1 line
73 KiB
JSON
1 line
73 KiB
JSON
{"repo": "6yte96/freebuffet", "files": {"provider-search.md": {"6f927e58de7c6cdba7e317904bf12b8798f59459": {"t": 1785595924.2668347, "n": 22594, "h": "546f7de87d9c5a91", "c": "## Directory of US and European API Providers\n\nThe following catalog lists US and European inference platforms, gateways, and aggregators [cite: 14, 23, 24, 26]. These services focus on reliable uptimes, edge optimization, and varied model selections [cite: 2, 14, 23].\n\n| Provider Name | Website URL | API Base URL (OpenAI-compatible) | Free Tier / Pricing Highlight | Notable Models Available | API Key Env Variable |\n| :--- | :--- | :--- | :--- | :--- | :--- |\n| **OpenRouter** | `https://openrouter.ai` | `https://openrouter.ai/api/v1` | 20+ free models, 20 RPM / 50 RPD limit [cite: 23] | DeepSeek R1, Llama 3.3, Gemma 3 [cite: 15, 27] | `OPENROUTER_API_KEY` |\n| **Google AI Studio** | `https://aistudio.google.com` | `https://generativelanguage.googleapis.com/v1beta` | 1500 RPD (Flash), 50 RPD (Pro) free [cite: 24] | Gemini 2.5 Flash, 1.5 Pro, Gemma 3 [cite: 23, 24] | `GEMINI_API_KEY` |\n| **Groq** | `https://groq.com` | `https://api.groq.com/openai/v1` | 30 RPM, 14,400 RPD permanent free tier [cite: 24, 28] | Llama 4 Scout, Llama 3.3 70B, Qwen3 [cite: 23, 25] | `GROQ_API_KEY` |\n| **Cloudflare Workers AI** | `https://cloudflare.com` | `https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/run` | 10K Neurons/day free (~300K/mo) [cite: 24, 29] | Llama 3.3 70B, Gemma 4, Kimi K2.5 [cite: 29] | `CLOUDFLARE_API_KEY` |\n| **GitHub Models** | `https://github.com` | `https://models.github.ai/inference` | Free prototyping for GitHub accounts [cite: 30] | GPT-5, Claude 3.5 Sonnet, DeepSeek R1 [cite: 28, 29] | `GITHUB_TOKEN` |\n| **Cohere** | `https://cohere.com` | `https://api.cohere.com/v2` | 1,000 requests/mo trial (non-commercial) [cite: 23, 31] | Command R+, Command A, Aya Expanse [cite: 22, 24] | `COHERE_API_KEY` |\n| **Mistral AI** | `https://mistral.ai` | `https://api.mistral.ai/v1` | Free Experiment Plan: 1B tokens/mo [cite: 20, 24] | Codestral, Mistral Large 3, Mistral Small 4 [cite: 21, 24] | `MISTRAL_API_KEY` |\n| **Cerebras** | `https://cerebras.ai` | `https://api.cerebras.ai/v1` | 1M tokens/day free [cite: 22, 23] | gpt-oss-120b, Llama 3.1 70B, zai-glm-4.7 [cite: 32] | `CEREBRAS_API_KEY` |\n| **Hugging Face** | `https://huggingface.co` | `https://router.huggingface.co/v1` | $0.10/mo serverless credits (renewed) [cite: 23, 31] | Llama 3.3 70B, Qwen 2.5 72B, Flux [cite: 23, 28] | `HF_TOKEN` |\n| **NVIDIA NIM** | `https://build.nvidia.com` | `https://integrate.api.nvidia.com/v1` | ~1,000 RPD, 40 RPM (phone verification) [cite: 20, 23] | Nemotron-3 Super, z-ai/glm-5.1, Kimi k2.6 [cite: 23, 26] | `NVIDIA_API_KEY` |\n| **SambaNova Cloud** | `https://sambanova.ai` | `https://api.sambanova.ai/v1` | Free tier (20-480 RPM), $5 trial credit [cite: 24, 33] | Llama 4 Maverick, DeepSeek R1, Llama 3.3 [cite: 33, 34] | `SAMBANOVA_API_KEY` |\n| **Together AI** | `https://together.ai` | `https://api.together.xyz/v1` | $1-5 signup credits, 200+ open-source models [cite: 33] | Llama 3.3, Qwen, DeepSeek [cite: 23, 33] | `TOGETHER_API_KEY` |\n| **Fireworks AI** | `https://fireworks.ai` | `https://api.fireworks.ai/inference/v1` | $1 trial credits (thousands of free calls) [cite: 22, 31] | Llama, Mixtral, Stable Diffusion [cite: 22, 31] | `FIREWORKS_API_KEY` |\n| **Baseten** | `https://baseten.co` | `https://app.baseten.co/v1` | $30 trial credit, pay by compute time [cite: 22, 31] | Serves deployable custom open models [cite: 14, 35] | `BASETEN_API_KEY` |\n| **Nebius AI** | `https://nebius.com` | `https://api.studio.nebius.com/v1` | $1 trial credits, EU-sovereign clouds [cite: 22, 31] | Qwen3-235b, Llama 3.3 [cite: 27, 31] | `NEBIUS_API_KEY` |\n| **DeepInfra** | `https://deepinfra.com` | `https://api.deepinfra.com/v1/openai` | $5 signup credits, low floor pricing ($0.02/1M) [cite: 23] | Qwen 3, DeepSeek V4, Llama 3.3 [cite: 36] | `DEEPINFRA_API_KEY` |\n| **Kimchi Inference** | `https://kimchi.dev` | `https://llm.cast.ai/openai/v1` | $50/mo free credits for developers [cite: 15, 37] | kimi-k2.6, minimax-m2.7, nemotron-3-super [cite: 38] | `KIMCHI_API_KEY` |\n| **TokenRouter** | `https://tokenrouter.io` | `https://api.tokenrouter.io/v1` | Free MiniMax M3 API access [cite: 17] | MiniMax M3, Claude 4, GPT-5 [cite: 3, 15] | `TOKENROUTER_API_KEY` |\n| **Aerolink** | `https://aerolink.ai` | `https://api.aerolink.ai/v1` | Free $350+ rolling credits ($10/5 hours) [cite: 15, 17] | Claude Opus, Claude Sonnet 4.6 [cite: 17] | `AEROLINK_API_KEY` |\n| **B.AI** | `https://b.ai` | `https://api.b.ai/v1` | 500K credits on signup, no card [cite: 15, 39] | Kimi k2.5, GLM-5, GLM 5.1 [cite: 39] | `BAI_API_KEY` |\n| **Lightning AI Models** | `https://lightning.ai` | `https://api.lightning.ai/v1` | 15 free credits/mo ($15 value, up to 30M tokens) [cite: 15, 16] | Claude, GPT, Gemini, DeepSeek [cite: 17, 39] | `LIGHTNING_API_KEY` |\n| **LLM7.io** | `https://llm7.io` | `https://api.llm7.io/v1` | 30 RPM basic tier without signup [cite: 22] | DeepSeek R1, Qwen2.5 Coder, Gemini 2.5 [cite: 29] | `LLM7_API_KEY` |\n| **Kluster AI** | `https://kluster.ai` | `https://api.kluster.ai/v1` | $5 trial credits, highly scalable [cite: 40] | DeepSeek R1, Llama 4, Qwen3-235B [cite: 22, 40] | `KLUSTER_API_KEY` |\n| **OVHcloud AI Endpoints** | `https://endpoints.ovh.com` | `https://oai.endpoints.kepler.ai.cloud.ovh.net/v1` | Free 2 RPM anonymous beta tier [cite: 26] | Qwen3Guard, gpt-oss-20b, Llama 3.3 [cite: 24, 26] | `OVH_API_KEY` |\n| **Glhf.chat** | `https://glhf.chat` | `https://glhf.chat/api/openai/v1` | Unlimited free models with high limits [cite: 23] | Llama 3.1 70B, Mixtral 8x7B [cite: 27] | `GLHF_API_KEY` |\n| **Aion Labs** | `https://aionlabs.ai` | `https://api.aionlabs.ai/v1` | Daily free token allowance on registration [cite: 30] | aion-2.0, aion-1.0, aion-1.0-mini [cite: 30] | `AION_API_KEY` |\n| **xAI (Grok)** | `https://x.ai` | `https://api.x.ai/v1` | $25 signup credits, no card required [cite: 30] | Grok-2, Grok-2 Mini [cite: 23] | `XAI_API_KEY` |\n| **Nscale** | `https://nscale.com` | `https://inference.api.nscale.com/v1` | Fair-use free tier on registration [cite: 27] | Llama 3.3 70B, DeepSeek R1 Distill [cite: 27] | `NSCALE_API_KEY` |\n| **AI21 Labs** | `https://ai21.com` | `https://api.ai21.com/studio/v1` | $10 trial credits expiring in 3 months [cite: 23, 29] | Jamba Large, Jamba Mini [cite: 30] | `AI21_API_KEY` |\n| **Upstage** | `https://upstage.ai` | `https://api.upstage.ai/v1` | $10 trial credits for Solar models [cite: 23, 31] | Solar Pro, Solar Mini [cite: 31] | `UPSTAGE_API_KEY` |\n| **NLP Cloud** | `https://nlpcloud.com` | `https://api.nlpcloud.com/v1` | $15 trial credits with phone verification [cite: 31] | Chat & translation models [cite: 18, 23] | `NLPCLOUD_API_KEY` |\n| **Modal** | `https://modal.com` | `https://api.modal.com` | $5/mo signup, $30/mo with card added [cite: 31] | Hostable open-weight models [cite: 31] | `MODAL_TOKEN_ID` |\n| **Inference.net** | `https://inference.net` | `https://api.inference.net/v1` | $1 free, $25 on survey completion [cite: 31] | DeepSeek R1, Llama 3.1 [cite: 23] | `INFERENCE_NET_API_KEY` |\n| **Hyperbolic** | `https://hyperbolic.xyz` | `https://api.hyperbolic.xyz/v1` | $1 free trial credit, 131K context [cite: 23, 40] | Llama 3.1 405B, DeepSeek V3 [cite: 23] | `HYPERBOLIC_API_KEY` |\n| **Scaleway** | `https://scaleway.com` | `https://api.scaleway.ai/v1` | 1M free tokens, EU/GDPR cloud [cite: 31, 41] | Mistral, Llama, Qwen [cite: 23] | `SCALEWAY_API_KEY` |\n| **OpenCode Zen** | `https://opencodezen.com` | `https://api.opencodezen.com/v1` | Free curated gateway [cite: 31] | Big Pickle Stealth, DeepSeek V4 Flash [cite: 31] | `OPENCODE_ZEN_API_KEY` |\n| **Replicate** | `https://replicate.com` | `https://api.replicate.com/v1` | Small trial credits at registration [cite: 23] | 1000+ text and generation models [cite: 14, 24] | `REPLICATE_API_TOKEN` |\n| **Lepton AI** | `https://lepton.ai` | `https://api.lepton.ai/api` | $10 free trial credits for developers [cite: 24, 40] | Llama, Mistral, Stable Diffusion [cite: 23] | `LEPTON_API_KEY` |\n| **Friendli AI** | `https://friendli.ai` | `https://api.friendli.ai/v1` | $10 free trial credits [cite: 23] | Popular open-weight models [cite: 23] | `FRIENDLI_API_KEY` |\n| **Cerebrium** | `https://cerebrium.com` | `https://api.cerebrium.com/v1` | $30 free trial credits [cite: 23] | Deployable open models, serverless [cite: 23, 37] | `CEREBRIUM_API_KEY` |\n| **Requesty** | `https://requesty.ai` | `https://api.requesty.ai/v1` | Free monthly credit allocations [cite: 23] | Multi-provider routing [cite: 23] | `REQUESTY_API_KEY` |\n| **Ollama Cloud** | `https://ollama.com` | `https://api.ollama.com` | Free usage-based tier resetting every 5 hours [cite: 26] | Llama, Qwen, Mistral [cite: 26] | `OLLAMA_API_KEY` |\n| **ZenMux** | `https://zenmux.com` | `https://api.zenmux.com/v1` | Free sandbox API tier [cite: 17] | Aggregated open-weight models [cite: 17] | `ZENMUX_API_KEY` |\n| **CrofAI** | `https://crofai.com` | `https://api.crofai.com/v1` | Promo credits at registration [cite: 26] | Cost-effective open weights [cite: 26] | `CROFAI_API_KEY` |\n| **Routeway AI** | `https://routeway.ai` | `https://api.routeway.ai/v1` | Free gateway routing tier [cite: 26] | Routed open-source models [cite: 26] | `ROUTEWAY_API_KEY` |\n| **FastRouter** | `https://fastrouter.ai` | `https://api.fastrouter.ai/v1` | Trial credits [cite: 26] | Multi-model routing [cite: 26] | `FASTROUTER_API_KEY` |\n| **AiQu** | `https://aiqu.se` | `https://api.aiqu.se/v1` | Swedish cloud free sandbox [cite: 14] | EU sovereign hosted open weights [cite: 37] | `AIQU_API_KEY` |\n| **Airon** | `https://airon.io` | `https://api.airon.io/v1` | Serverless trial credits [cite: 37] | Hosted open-source LLMs [cite: 37] | `AIRON_API_KEY` |\n| **AKI.IO** | `https://aki.io` | `https://api.aki.io/v1` | 10 EUR trial credits [cite: 37] | Llama, Qwen, Mistral [cite: 37] | `AKI_API_KEY` |\n| **ARK Labs** | `https://arklabs.ai` | `https://api.arklabs.ai/v1` | EU sovereign free tier [cite: 37] | Regulated environment models [cite: 37] | `ARK_API_KEY` |\n| **Beam** | `https://beam.cloud` | `https://api.beam.cloud/v1` | Serverless GPU free trial tier [cite: 37] | Custom hostable models [cite: 37] | `BEAM_API_KEY` |\n| **BentoCloud** | `https://bentoml.com` | `https://api.bentocloud.com/v1` | Free trial credits [cite: 37] | Llama 2, SD, Flan-T5 [cite: 37] | `BENTOML_API_KEY` |\n| **Berget AI** | `https://berget.ai` | `https://api.berget.ai/v1` | EU sovereign free tier [cite: 37] | EU compliant models [cite: 37] | `BERGET_API_KEY` |\n| **Cortecs AI** | `https://cortecs.ai` | `https://api.cortecs.ai/v1` | Gateway free trial [cite: 37] | European smart-routing models [cite: 37] | `CORTECS_API_KEY` |\n| **EUrouter** | `https://eurouter.com` | `https://api.eurouter.com/v1` | 1K requests/month free [cite: 37] | Over 100 GDPR models [cite: 37] | `EUROUTER_API_KEY` |\n| **Infercom** | `https://infercom.ai` | `https://api.infercom.ai/v1` | Free tier, EU hosted [cite: 37] | Sovereign open models [cite: 37] | `INFERCOM_API_KEY` |\n| **IonRouter** | `https://ionrouter.com` | `https://api.ionrouter.com/v1` | Free tier options [cite: 37] | Open-source LLMs [cite: 37] | `IONROUTER_API_KEY` |\n| **Jina AI** | `https://jina.ai` | `https://api.jina.ai/v1` | 10M free tokens on signup [cite: 37] | Embeddings & rerankers [cite: 37] | `JINA_API_KEY` |\n| **Lambda AI** | `https://lambdalabs.com` | `https://api.lambdalabs.com/v1` | Trial cloud credits [cite: 14, 40] | Deployable LLMs [cite: 37] | `LAMBDA_API_KEY` |\n| **Monster API** | `https://monsterapi.ai` | `https://api.monsterapi.ai/v1` | Free trial credits [cite: 37] | Llama 2, Zephyr, Falcon [cite: 37] | `MONSTER_API_KEY` |\n| **OctoAI** | `https://octoai.cloud` | `https://api.octoai.cloud/v1` | Free trial credits [cite: 37] | Production GenAI models [cite: 37] | `OCTOAI_API_KEY` |\n| **Prem AI** | `https://premai.io` | `https://api.premai.io/v1` | Free tier [cite: 37] | Unified deployment models [cite: 37] | `PREM_API_KEY` |\n| **Synexa** | `https://synexa.io` | `https://api.synexa.io/v1` | $0.0015/image, trial credits [cite: 37] | 100+ production models [cite: 37] | `SYNEXA_API_KEY` |\n| **Tokenware** | `https://tokenware.io` | `https://api.tokenware.io/v1` | Free trial credits [cite: 37] | 200+ models [cite: 37] | `TOKENWARE_API_KEY` |\n| **Portkey AI** | `https://portkey.ai` | `https://api.portkey.sh/v1` | Free developer tier with request telemetry [cite: 2] | Provider models mapped dynamically [cite: 1] | `PORTKEY_API_KEY` |\n| **Voyage AI** | `https://voyageai.com` | `https://api.voyageai.com/v1` | Free tier credits available [cite: 37] | High quality embeddings [cite: 37] | `VOYAGE_API_KEY` |\n| **Featherless AI**| `https://featherless.ai` | `https://api.featherless.ai/v1` | Free tier trial [cite: 41] | 200+ open-source models [cite: 41] | `FEATHERLESS_API_KEY` |\n| **Morph** | `https://morph.chat` | `https://api.morph.chat/v1` | Agentic free credits [cite: 41] | Autonomous LLMs [cite: 41] | `MORPH_API_KEY` |\n| **Predibase** | `https://predibase.com` | `https://connect.predibase.com/v1` | Free credits on signup [cite: 41] | Custom fine-tuned LoRA models [cite: 41] | `PREDIBASE_API_KEY` |\n| **FerryAPI** | `https://ferryapi.com` | `https://api.ferryapi.com/v1` | Prepay signup bonus [cite: 37] | Qwen, DeepSeek, Kimi [cite: 37] | `FERRYAPI_API_KEY` |\n| **Fast Pivot** | `https://fastpivot.io` | `https://api.fastpivot.io/v1` | Prepaid credit trial [cite: 37] | Aggregated open-weight models [cite: 37] | `FASTPIVOT_API_KEY` |\n| **General Compute**| `https://generalcompute.com`| `https://api.generalcompute.com/v1`| Trial compute hours [cite: 37] | Latency-critical coding models [cite: 37] | `GENERAL_COMPUTE_API_KEY` |\n| **CoreWeave** | `https://coreweave.com` | `https://api.coreweave.com/v1` | Trial cloud credits [cite: 37] | Deployable open LLMs [cite: 37] | `COREWEAVE_API_KEY` |\n| **Amazon Bedrock**| `https://aws.amazon.com/bedrock`| `https://bedrock-runtime.{region}.amazonaws.com`| AWS Free Tier credits [cite: 37] | Claude, Llama, Jurassic [cite: 37] | `AWS_SECRET_ACCESS_KEY` |\n| **Anyscale** | `https://anyscale.com` | `https://api.endpoints.anyscale.com/v1`| Signup trial credits [cite: 37] | Curated open weight models [cite: 37] | `ANYSCALE_API_KEY` |\n| **DeepInfra Native**| `https://deepinfra.com` | `https://api.deepinfra.com/v1/inference/{model}` | $5 Native API credits [cite: 42, 43] | Flux, Speech, zero-shot models [cite: 42] | `DEEPINFRA_API_KEY` |\n| **Perplexity API** | `https://perplexity.ai` | `https://api.perplexity.ai` | Trial credits for API subscribers [cite: 44] | Llama 3.1 70B, Qwen 2.5 72B [cite: 26] | `PERPLEXITY_API_KEY` |\n| **Venice.ai** | `https://venice.ai` | `https://api.venice.ai/v1` | Daily limits allowed for free tier [cite: 23] | Llama 3.1 405B, Dolphin Mixtral [cite: 23] | `VENICE_API_KEY` |\n| **Chutes.ai** | `https://chutes.ai` | `https://api.chutes.ai/v1` | Free community tier [cite: 23] | DeepSeek-R1, Qwen 2.5 72B [cite: 23] | `CHUTES_API_KEY` |\n| **Packet.ai** | `https://packet.ai` | `https://api.packet.ai/v1` | Billed per token, trial credits [cite: 37] | RTX PRO 6000 managed models [cite: 37] | `PACKET_API_KEY` |\n| **Verda** | `https://verda.ai` | `https://api.verda.ai/v1` | Serverless trial [cite: 37] | European optimized LLMs [cite: 37] | `VERDA_API_KEY` |\n\n---\n\n## Directory of Chinese and Asian API Providers\n\nProviders within this region often deliver highly aggressive per-token pricing structures [cite: 10]. These infrastructures are heavily optimized for both dense and mixture-of-experts open weights [cite: 9, 10].\n\n| Provider Name | Website URL | API Base URL (OpenAI-compatible) | Free Tier / Pricing Highlight | Notable Models Available | API Key Env Variable |\n| :--- | :--- | :--- | :--- | :--- | :--- |\n| **SiliconFlow** | `https://siliconflow.cn` | `https://api.siliconflow.cn/v1` | Permanent free models, 1K RPM [cite: 26, 40] | DeepSeek R1, MiniMax-M2, Qwen3 [cite: 10, 14, 26] | `SILICONFLOW_API_KEY` |\n| **Z AI (Zhipu AI)** | `https://bigmodel.cn` | `https://open.bigmodel.cn/api/paas/v4` | Permanent free models, no card required [cite: 30] | GLM-4.7-Flash, GLM-4.5-Flash [cite: 30] | `ZHIPU_API_KEY` |\n| **ModelScope** | `https://modelscope.cn` | `https://api-inference.modelscope.cn/v1` | 2,000 requests/day free, registration only [cite: 27] | Qwen 3.5, Qwen Image [cite: 27] | `MODELSCOPE_API_KEY` |\n| **Alibaba Model Studio**| `https://alibabacloud.com` | `https://dashscope-intl.aliyuncs.com/compatible-mode/v1` | 1M free tokens per Qwen model [cite: 30] | Qwen series, open weight models [cite: 26, 29] | `DASHSCOPE_API_KEY` |\n| **DeepSeek AI** | `https://deepseek.com` | `https://api.deepseek.com/v1` | 10M free tokens on signup, cheap pricing [cite: 23] | DeepSeek-V3, DeepSeek-R1 [cite: 23] | `DEEPSEEK_API_KEY` |\n| **MiniMax AI** | `https://minimaxi.com` | `https://api.minimax.chat/v1` | Free trial credits on registration [cite: 17] | MiniMax M3, M2.7 [cite: 17] | `MINIMAX_API_KEY` |\n| **Moonshot AI / Kimi**| `https://kimi.ai` | `https://api.moonshot.cn/v1` | Free trial credits at signup [cite: 31] | Kimi-K2.7, Kimi-VL [cite: 15, 45] | `MOONSHOT_API_KEY` |\n| **Yi AI (01.AI)** | `https://01.ai` | `https://api.01.ai/v1` | Initial trial credits on signup [cite: 23] | Yi-Large, Yi-34B [cite: 23] | `YI_API_KEY` |\n| **Tencent Hunyuan**| `https://cloud.tencent.com`| `https://api.hunyuan.tencent.com/v1` | Free trial credits for cloud developers [cite: 25] | Hunyuan-Turbo, Hunyuan-Pro [cite: 20] | `HUNYUAN_API_KEY` |\n| **Baidu Qianfan** | `https://cloud.baidu.com` | `https://qianfan.baidubce.com/v2` | Free trial credits, ERNIE Speed free tier [cite: 20] | ERNIE-Lite, ERNIE-Speed [cite: 20] | `QIANFAN_API_KEY` |\n| **ByteDance Ark** | `https://volcengine.com` | `https://ark.cn-beijing.volces.com/api/v3` | Millions of free signup tokens [cite: 20] | Doubao-pro, Doubao-lite [cite: 20] | `VOLCENGINE_API_KEY` |\n| **Spark AI (iFlytek)**| `https://xfyun.cn` | `https://spark-api.xf-yun.com/v1/chat` | 2M-5M free tokens on signup [cite: 20] | Spark-Pro, Spark-Lite [cite: 20] | `SPARK_API_KEY` |\n| **Xiaomi Mimo** | `https://mi.com` | `https://api.mimo.xiaomi.com/v1` | Highly competitive pricing ($0.14/1M) [cite: 41] | mimo-v2.5-pro, mimo-v2.5 [cite: 27] | `MIMO_API_KEY` |\n| **StreamLake (Kuaishou)**| `https://streamlake.com`| `https://api.streamlake.com/v1` | Developer trial credits [cite: 20] | Kuaishou KwaiYong [cite: 20] | `STREAMLAKE_API_KEY` |\n| **Stepfun** | `https://stepfun.com` | `https://api.stepfun.com/v1` | Registration free credits [cite: 20] | Step-3.5-flash, Step-3-pro [cite: 20] | `STEPFUN_API_KEY` |\n| **Gitee AI** | `https://gitee.com/ai` | `https://ai.gitee.com/v1` | Developer free sandbox tokens [cite: 20] | Hosted Qwen and DeepSeek models [cite: 20] | `GITEE_AI_TOKEN` |\n| **Baichuan AI** | `https://baichuan-ai.com` | `https://api.baichuan-ai.com/v1` | Trial credits on signup [cite: 20] | Baichuan4, Baichuan2 [cite: 20] | `BAICHUAN_API_KEY` |\n| **SenseTime** | `https://sensetime.com` | `https://api.sensenova.cn/v1` | Trial credits for new developers [cite: 20] | SenseChat-5, SenseChat-4 [cite: 20] | `SENSENOVA_API_KEY` |\n| **Linkos AI** | `https://linkos.ai` | `https://api.linkos.ai/v1` | Gateway free trial [cite: 20] | Routed Asian frontier models [cite: 20] | `LINKOS_API_KEY` |\n| **Baobaogou AI** | `https://baobaogou.com` | `https://api.baobaogou.com/v1` | Regional trial credits [cite: 20] | Open-source models [cite: 20] | `BAOBAOGOU_API_KEY` |\n\n---\n\n## Directory of Local, Self-Hosted, and Gateway Solutions\n\nLocal engines provide absolute data privacy and zero token-based operational expenditure after hardware provisioning [cite: 21, 22]. Local proxies act as compliant boundary nodes [cite: 21].\n\n| Platform Name | Project URL | Local API Base URL | Cost Highlight | Notable Models Serviced | Key Implementation Command / Env |\n| :--- | :--- | :--- | :--- | :--- | :--- |\n| **AegisGate** | `https://github.com/ax128/AegisGate` | `http://localhost:18080/v1` | $0, self-hosted security gateway [cite: 21] | Proxies Claude, GPT, Llama [cite: 21] | `docker compose up -d` [cite: 21] |\n| **FreeLLMAPI** | `https://github.com/open-free-llm-api/freellm-hub-data` | `http://localhost:3000/v1` | $0, self-hosted proxy aggregator [cite: 37] | Routes 60+ models [cite: 46] | `git clone` + configure keys [cite: 46] |\n| **LM Studio** | `https://lmstudio.ai` | `http://localhost:1234/v1` | Free local GUI & API engine [cite: 20] | Llama, Qwen GGUF formats [cite: 20] | Launch Local Server in UI [cite: 20] |\n| **Ollama Local** | `https://ollama.com` | `http://localhost:11434/v1` | Free local command-line engine [cite: 20] | Llama 3, Qwen 2.5, Mistral [cite: 20] | `ollama run llama3` |\n| **llama.cpp** | `https://github.com/ggerganov/llama.cpp` | `http://localhost:8080/v1` | Free local C/C++ inference engine [cite: 20] | Any GGUF models [cite: 20] | `./llama-server` [cite: 20] |\n| **vLLM Engine** | `https://github.com/vllm-project/vllm` | `http://localhost:8000/v1` | Free high-throughput local engine [cite: 2, 4] | AWQ, GPTQ, FP16 serverless [cite: 4] | `python -m vllm.entrypoints.openai.api_server` |\n| **SGLang Server** | `https://github.com/sgl-project/sglang` | `http://localhost:30000/v1` | Free latency-optimized local server [cite: 37] | High-throughput open weights [cite: 37] | `python -m sglang.launch_server` |\n| **LocalAI** | `https://github.com/mudler/LocalAI` | `http://localhost:8080/v1` | Free multi-modality local server | Llama, Whisper, Stable Diffusion | `docker run localai` |\n| **Tabby AI** | `https://github.com/TabbyML/tabby` | `http://localhost:8080/v1` | Free self-hosted code completion engine | Tabby coder models | `tabby serve --model` |\n| **Text Gen WebUI**| `https://github.com/oobabooga/text-generation-webui` | `http://localhost:5000/v1` | Free local web browser interface | Transformers, ExLlamaV2 | `python server.py --api` |\n| **KoboldCPP** | `https://github.com/LostRuins/koboldcpp` | `http://localhost:5001/v1` | Free local GGUF parser engine | GGUF text models | `python koboldcpp.py` |\n| **Aphrodite** | `https://github.com/PygmalionAI/aphrodite-engine` | `http://localhost:2242/v1` | Free Pygmalion local inference | Pygmalion / Hugging Face weights | `python -m aphrodite.endpoints.openai.api_server` |\n| **MindSpore** | `https://github.com/mindspore-ai/mindspore` | `http://localhost:8080/v1` | Free local deep learning framework | MindSpore-optimized models | `python -m mindspore.entrypoints` |\n| **ONNX Runtime** | `https://github.com/microsoft/onnxruntime` | `http://localhost:8090/v1` | Free multi-platform hardware acceleration | ONNX quantized text models | Launch ONNX local API endpoint |\n\n"}}, "src/providers.ts": {"6f927e58de7c6cdba7e317904bf12b8798f59459": {"t": 1785595924.4037778, "n": 47621, "h": "6203e2d9d695093e", "c": "export interface Provider {\n id: string\n name: string\n baseURL: string\n envKey?: string\n website: string\n apiType: \"openai\" | \"anthropic\" | \"openai-compatible\"\n models?: string[]\n description?: string\n freeTier?: string\n noCc?: boolean\n permanentFree?: boolean\n recommended?: boolean\n tags?: string[]\n}\n\nconst PROVIDERS: Provider[] = [\n {\n id: \"groq\",\n name: \"Groq\",\n baseURL: \"https://api.groq.com/openai/v1\",\n envKey: \"GROQ_API_KEY\",\n website: \"https://console.groq.com\",\n apiType: \"openai-compatible\",\n freeTier: \"30 RPM, 1000-14400 RPD, no CC\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"cerebras\",\n name: \"Cerebras\",\n baseURL: \"https://api.cerebras.ai/v1\",\n envKey: \"CEREBRAS_API_KEY\",\n website: \"https://cloud.cerebras.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"30 RPM, 1M tok/day, no CC, 2600 tok/s\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"google-gemini\",\n name: \"Google Gemini\",\n baseURL: \"https://generativelanguage.googleapis.com/v1beta\",\n envKey: \"GOOGLE_API_KEY\",\n website: \"https://aistudio.google.com\",\n apiType: \"openai-compatible\",\n freeTier: \"15-30 RPM, 50-1500 RPD, no CC\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"github-models\",\n name: \"GitHub Models\",\n baseURL: \"https://models.github.ai/inference\",\n envKey: \"GITHUB_TOKEN\",\n website: \"https://github.com/marketplace/models\",\n apiType: \"openai-compatible\",\n freeTier: \"10-15 RPM, 50-150 RPD, needs GitHub account\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"mistral\",\n name: \"Mistral AI (La Plateforme)\",\n baseURL: \"https://api.mistral.ai/v1\",\n envKey: \"MISTRAL_API_KEY\",\n website: \"https://console.mistral.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"~1 RPS, ~500K TPM, 1B tok/month, no CC\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"cloudflare-workers-ai\",\n name: \"Cloudflare Workers AI\",\n baseURL: \"https://api.cloudflare.com/client/v4/accounts/{account}/ai/v1\",\n envKey: \"CLOUDFLARE_API_TOKEN\",\n website: \"https://dash.cloudflare.com\",\n apiType: \"openai-compatible\",\n freeTier: \"10K neurons/day, 300 RPM, no CC\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"huggingface\",\n name: \"Hugging Face Inference\",\n baseURL: \"https://router.huggingface.co/v1\",\n envKey: \"HF_TOKEN\",\n website: \"https://huggingface.co/settings/tokens\",\n apiType: \"openai-compatible\",\n freeTier: \"100K credits/month, no CC\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"zhipu\",\n name: \"Zhipu AI / GLM (智谱)\",\n baseURL: \"https://open.bigmodel.cn/api/paas/v4\",\n envKey: \"ZHIPU_API_KEY\",\n website: \"https://open.bigmodel.cn\",\n apiType: \"openai-compatible\",\n freeTier: \"Free Flash models, no CC\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"ovhcloud\",\n name: \"OVHcloud AI Endpoints\",\n baseURL: \"https://oai.endpoints.kepler.ai.cloud.ovh.net/v1\",\n envKey: \"OVH_AI_ENDPOINTS_ACCESS_TOKEN\",\n website: \"https://www.ovhcloud.com/en/public-cloud/ai-endpoints\",\n apiType: \"openai-compatible\",\n freeTier: \"2 RPM/IP anonymous, 400 RPM with key, $200 trial\",\n noCc: true,\n permanentFree: true,\n description: \"Hosted in EU, GDPR-compliant\",\n },\n {\n id: \"nvidia-nim\",\n name: \"NVIDIA NIM\",\n baseURL: \"https://integrate.api.nvidia.com/v1\",\n envKey: \"NVIDIA_API_KEY\",\n website: \"https://build.nvidia.com\",\n apiType: \"openai-compatible\",\n freeTier: \"~40 RPM, no daily cap, phone verification\",\n permanentFree: true,\n description: \"75+ free models\",\n },\n {\n id: \"codestral\",\n name: \"Codestral (Mistral Code)\",\n baseURL: \"https://api.mistral.ai/v1\",\n envKey: \"MISTRAL_API_KEY\",\n website: \"https://console.mistral.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"30 RPM, 2K req/day, phone verification\",\n description: \"256K context, code-focused\",\n },\n {\n id: \"llm7\",\n name: \"LLM7.io\",\n baseURL: \"https://api.llm7.io/v1\",\n envKey: \"LLM7_API_KEY\",\n website: \"https://llm7.io\",\n apiType: \"openai-compatible\",\n freeTier: \"30 RPM (120 RPM with free token), no CC\",\n noCc: true,\n permanentFree: true,\n description: \"GDPR-compliant\",\n },\n {\n id: \"kilo-code\",\n name: \"Kilo Code\",\n baseURL: \"https://api.kilo.ai/api/gateway\",\n envKey: \"KILO_API_KEY\",\n website: \"https://kilo.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"~200 req/hr, free auto-router\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"openrouter\",\n name: \"OpenRouter\",\n baseURL: \"https://openrouter.ai/api/v1\",\n envKey: \"OPENROUTER_API_KEY\",\n website: \"https://openrouter.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"50-200 RPD on :free models, 22+ free models\",\n permanentFree: true,\n },\n {\n id: \"sambanova\",\n name: \"SambaNova\",\n baseURL: \"https://api.sambanova.ai/v1\",\n envKey: \"SAMBANOVA_API_KEY\",\n website: \"https://cloud.sambanova.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"20 RPM, 20 RPD, 200K TPD + $5 credits\",\n permanentFree: true,\n },\n {\n id: \"cohere\",\n name: \"Cohere\",\n baseURL: \"https://api.cohere.com/v2\",\n envKey: \"COHERE_API_KEY\",\n website: \"https://dashboard.cohere.com\",\n apiType: \"openai-compatible\",\n freeTier: \"1000 calls/month Trial, 20 RPM, no CC\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"ai21\",\n name: \"AI21 Labs\",\n baseURL: \"https://api.ai21.com/studio/v1\",\n envKey: \"AI21_API_KEY\",\n website: \"https://studio.ai21.com\",\n apiType: \"openai-compatible\",\n freeTier: \"200 RPM, 10 RPS, no CC\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"siliconflow\",\n name: \"SiliconFlow (硅基流动)\",\n baseURL: \"https://api.siliconflow.cn/v1\",\n envKey: \"SILICONFLOW_API_KEY\",\n website: \"https://siliconflow.cn\",\n apiType: \"openai-compatible\",\n freeTier: \"30 RPM, 60K TPM on free models\",\n permanentFree: true,\n },\n {\n id: \"modelscope\",\n name: \"ModelScope (魔搭)\",\n baseURL: \"https://api-inference.modelscope.cn/v1\",\n envKey: \"MODELSCOPE_API_KEY\",\n website: \"https://modelscope.cn\",\n apiType: \"openai-compatible\",\n freeTier: \"2000 RPD, ≤500 RPD/model\",\n permanentFree: true,\n },\n {\n id: \"deepseek\",\n name: \"DeepSeek\",\n baseURL: \"https://api.deepseek.com/v1\",\n envKey: \"DEEPSEEK_API_KEY\",\n website: \"https://platform.deepseek.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Dynamic rate limits\",\n permanentFree: true,\n },\n {\n id: \"glhf\",\n name: \"Glhf.chat\",\n baseURL: \"https://glhf.chat/api/openai/v1\",\n envKey: \"GLHF_API_KEY\",\n website: \"https://glhf.chat\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited for free models\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"chutes\",\n name: \"Chutes.ai\",\n baseURL: \"https://api.chutes.ai/v1\",\n envKey: \"CHUTES_API_KEY\",\n website: \"https://chutes.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Community-powered, no hard cap\",\n permanentFree: true,\n },\n {\n id: \"kluster\",\n name: \"Kluster.ai\",\n baseURL: \"https://api.kluster.ai/v1\",\n envKey: \"KLUSTER_API_KEY\",\n website: \"https://kluster.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Batch-based, free for API\",\n permanentFree: true,\n description: \"Llama 3.1 405B free tier\",\n },\n {\n id: \"nscale\",\n name: \"Nscale\",\n baseURL: \"https://inference.api.nscale.com/v1\",\n envKey: \"NSCALE_API_KEY\",\n website: \"https://console.nscale.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Fair-use free tier\",\n permanentFree: true,\n },\n {\n id: \"inference-net\",\n name: \"Inference.net\",\n baseURL: \"https://api.inference.net/v1\",\n envKey: \"INFERENCE_NET_API_KEY\",\n website: \"https://inference.net\",\n apiType: \"openai-compatible\",\n freeTier: \"Free for listed models\",\n permanentFree: true,\n },\n {\n id: \"pollinations\",\n name: \"Pollinations AI\",\n baseURL: \"https://gen.pollinations.ai\",\n website: \"https://pollinations.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Free, no API key needed\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"puter\",\n name: \"Puter\",\n baseURL: \"https://puter.com/api/llm\",\n website: \"https://puter.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free, no signup\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"scaleway\",\n name: \"Scaleway Generative APIs\",\n baseURL: \"https://api.scaleway.ai/v1\",\n envKey: \"SCALEWAY_API_KEY\",\n website: \"https://console.scaleway.com\",\n apiType: \"openai-compatible\",\n freeTier: \"1M free tokens\",\n permanentFree: true,\n },\n {\n id: \"nebulous\",\n name: \"Nebius AI Studio\",\n baseURL: \"https://api.studio.nebius.ai/v1\",\n envKey: \"NEBIUS_API_KEY\",\n website: \"https://studio.nebius.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Tier-based free access\",\n permanentFree: true,\n },\n {\n id: \"openai\",\n name: \"OpenAI\",\n baseURL: \"https://api.openai.com/v1\",\n envKey: \"OPENAI_API_KEY\",\n website: \"https://platform.openai.com\",\n apiType: \"openai\",\n freeTier: \"$5 free credits (90 day expiry)\",\n },\n {\n id: \"anthropic\",\n name: \"Anthropic\",\n baseURL: \"https://api.anthropic.com/v1\",\n envKey: \"ANTHROPIC_API_KEY\",\n website: \"https://console.anthropic.com\",\n apiType: \"anthropic\",\n freeTier: \"Usage credits with Claude subscription\",\n },\n {\n id: \"xai\",\n name: \"xAI (Grok)\",\n baseURL: \"https://api.x.ai/v1\",\n envKey: \"XAI_API_KEY\",\n website: \"https://console.x.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"$25/month free credits for Grok-2\",\n description: \"2M context\",\n },\n {\n id: \"together\",\n name: \"Together AI\",\n baseURL: \"https://api.together.xyz/v1\",\n envKey: \"TOGETHER_API_KEY\",\n website: \"https://together.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"$5 free credits, 100+ models\",\n },\n {\n id: \"fireworks\",\n name: \"Fireworks AI\",\n baseURL: \"https://api.fireworks.ai/inference/v1\",\n envKey: \"FIREWORKS_API_KEY\",\n website: \"https://fireworks.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"$1 free credits, 10 RPM without payment\",\n },\n {\n id: \"deepinfra\",\n name: \"Deep Infra\",\n baseURL: \"https://api.deepinfra.com/v1/openai\",\n envKey: \"DEEPINFRA_API_KEY\",\n website: \"https://deepinfra.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free credits on signup, 200 concurrent requests\",\n },\n {\n id: \"hyperbolic\",\n name: \"Hyperbolic\",\n baseURL: \"https://api.hyperbolic.xyz/v1\",\n envKey: \"HYPERBOLIC_API_KEY\",\n website: \"https://app.hyperbolic.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"$1 free credits (permanent)\",\n permanentFree: true,\n },\n {\n id: \"novita\",\n name: \"Novita AI\",\n baseURL: \"https://api.novita.ai/openai\",\n envKey: \"NOVITA_API_KEY\",\n website: \"https://novita.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"$0.50 free credits/month\",\n },\n {\n id: \"dashscope\",\n name: \"DashScope / Alibaba (灵积)\",\n baseURL: \"https://dashscope-intl.aliyuncs.com/compatible-mode/v1\",\n envKey: \"DASHSCOPE_API_KEY\",\n website: \"https://bailian.console.alibabacloud.com\",\n apiType: \"openai-compatible\",\n freeTier: \"1M free tokens per model (90 days)\",\n },\n {\n id: \"modal\",\n name: \"Modal\",\n baseURL: \"https://api.modal.com/v1\",\n envKey: \"MODAL_API_KEY\",\n website: \"https://modal.com\",\n apiType: \"openai-compatible\",\n freeTier: \"$30/month free credits\",\n },\n {\n id: \"replicate\",\n name: \"Replicate\",\n baseURL: \"https://api.replicate.com/v1\",\n envKey: \"REPLICATE_API_TOKEN\",\n website: \"https://replicate.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Development quota\",\n },\n {\n id: \"perplexity\",\n name: \"Perplexity AI\",\n baseURL: \"https://api.perplexity.ai/v1\",\n envKey: \"PERPLEXITY_API_KEY\",\n website: \"https://www.perplexity.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"~50 RPM free tier\",\n },\n {\n id: \"baseten\",\n name: \"Baseten\",\n baseURL: \"https://app.baseten.co/v1\",\n envKey: \"BASETEN_API_KEY\",\n website: \"https://baseten.co\",\n apiType: \"openai-compatible\",\n freeTier: \"Trial credits available\",\n },\n {\n id: \"upstage\",\n name: \"Upstage AI (Solar)\",\n baseURL: \"https://api.upstage.ai/v1/solar\",\n envKey: \"UPSTAGE_API_KEY\",\n website: \"https://upstage.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Trial credits available\",\n },\n {\n id: \"aiml-api\",\n name: \"AIML API\",\n baseURL: \"https://api.aimlapi.com/v1\",\n envKey: \"AIML_API_KEY\",\n website: \"https://aimlapi.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free tier, 200+ models\",\n },\n {\n id: \"coze\",\n name: \"Coze\",\n baseURL: \"https://api.coze.com/v1\",\n envKey: \"COZE_API_KEY\",\n website: \"https://coze.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free daily tokens\",\n },\n {\n id: \"moonshot\",\n name: \"Moonshot AI / Kimi (月之暗面)\",\n baseURL: \"https://api.moonshot.cn/v1\",\n envKey: \"MOONSHOT_API_KEY\",\n website: \"https://platform.moonshot.cn\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"minimax\",\n name: \"MiniMax (稀宇)\",\n baseURL: \"https://api.minimax.chat/v1\",\n envKey: \"MINIMAX_API_KEY\",\n website: \"https://platform.minimax.chat\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"baichuan\",\n name: \"Baichuan AI (百川)\",\n baseURL: \"https://api.baichuan-ai.com/v1\",\n envKey: \"BAICHUAN_API_KEY\",\n website: \"https://platform.baichuan-ai.com\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"stepfun\",\n name: \"Stepfun / Step (阶跃星辰)\",\n baseURL: \"https://api.stepfun.com/v1\",\n envKey: \"STEPFUN_API_KEY\",\n website: \"https://platform.stepfun.com\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"lingyi\",\n name: \"Lingyiwanwu / 01.AI (零一万物)\",\n baseURL: \"https://api.lingyiwanwu.com/v1\",\n envKey: \"LINGYI_API_KEY\",\n website: \"https://platform.lingyiwanwu.com\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"baidu\",\n name: \"Baidu ERNIE (文心)\",\n baseURL: \"https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat\",\n envKey: \"BAIDU_API_KEY\",\n website: \"https://cloud.baidu.com\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"tencent\",\n name: \"Tencent Hunyuan (混元)\",\n baseURL: \"https://api.hunyuan.cloud.tencent.com/v1\",\n envKey: \"TENCENT_API_KEY\",\n website: \"https://hunyuan.tencent.com\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"volcengine\",\n name: \"ByteDance Volcengine / Ark (火山引擎)\",\n baseURL: \"https://ark.cn-beijing.volces.com/api/v3\",\n envKey: \"VOLC_API_KEY\",\n website: \"https://console.volcengine.com/ark\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"sensetime\",\n name: \"SenseTime / SenseNova (商汤)\",\n baseURL: \"https://api.sensenova.cn/v1\",\n envKey: \"SENSENOVA_API_KEY\",\n website: \"https://platform.sensenova.cn\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"baidu-qianfan\",\n name: \"Baidu Qianfan (千帆)\",\n baseURL: \"https://qianfan.baidubce.com/v2\",\n envKey: \"QIANFAN_API_KEY\",\n website: \"https://console.bce.baidu.com/qianfan\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"iflytek\",\n name: \"iFlytek Spark (讯飞星火)\",\n baseURL: \"https://spark-api.xf-yun.com/v3.5/chat\",\n envKey: \"IFLYTEK_API_KEY\",\n website: \"https://xinghuo.xfyun.cn\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"azure-openai\",\n name: \"Azure OpenAI\",\n baseURL: \"https://{resource}.openai.azure.com/openai\",\n envKey: \"AZURE_OPENAI_API_KEY\",\n website: \"https://portal.azure.com\",\n apiType: \"openai\",\n freeTier: \"$200 Azure free credits (30 days)\",\n },\n {\n id: \"aws-bedrock\",\n name: \"Amazon Bedrock\",\n baseURL: \"https://bedrock-runtime.{region}.amazonaws.com\",\n envKey: \"AWS_ACCESS_KEY_ID\",\n website: \"https://aws.amazon.com/bedrock\",\n apiType: \"openai-compatible\",\n freeTier: \"Free tier with AWS Free Tier\",\n },\n {\n id: \"gcp-vertex\",\n name: \"Google Vertex AI\",\n baseURL: \"https://{region}-aiplatform.googleapis.com/v1\",\n envKey: \"GOOGLE_APPLICATION_CREDENTIALS\",\n website: \"https://console.cloud.google.com/vertex-ai\",\n apiType: \"openai-compatible\",\n freeTier: \"$300 Google Cloud free credits\",\n },\n {\n id: \"cloudflare-gateway\",\n name: \"Cloudflare AI Gateway\",\n baseURL: \"https://gateway.ai.cloudflare.com/v1/{account}/{gateway}\",\n envKey: \"CLOUDFLARE_API_TOKEN\",\n website: \"https://dash.cloudflare.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free tier with Cloudflare account\",\n },\n {\n id: \"vercel-gateway\",\n name: \"Vercel AI Gateway\",\n baseURL: \"https://gateway.vercel.ai/v1\",\n envKey: \"VERCEL_API_TOKEN\",\n website: \"https://vercel.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free tier through Vercel\",\n },\n {\n id: \"digitalocean\",\n name: \"DigitalOcean Inference\",\n baseURL: \"https://{region}.inference.digitalocean.com/v1\",\n envKey: \"DIGITALOCEAN_ACCESS_TOKEN\",\n website: \"https://cloud.digitalocean.com\",\n apiType: \"openai-compatible\",\n freeTier: \"$200 free credits (60 days)\",\n },\n {\n id: \"oci\",\n name: \"Oracle Cloud OCI Generative AI\",\n baseURL: \"https://inference.generativeai.{region}.oci.oraclecloud.com/v1\",\n envKey: \"OCI_API_KEY\",\n website: \"https://cloud.oracle.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Always Free tier + $300 trial\",\n },\n {\n id: \"helicone\",\n name: \"Helicone AI Gateway\",\n baseURL: \"https://ai-gateway.helicone.ai/v1\",\n envKey: \"HELICONE_API_KEY\",\n website: \"https://helicone.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"100K requests/month free\",\n permanentFree: true,\n },\n {\n id: \"portkey\",\n name: \"Portkey AI Gateway\",\n baseURL: \"https://api.portkey.ai/v1\",\n envKey: \"PORTKEY_API_KEY\",\n website: \"https://portkey.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"10K requests/month free\",\n permanentFree: true,\n },\n {\n id: \"litellm\",\n name: \"LiteLLM Proxy\",\n baseURL: \"http://localhost:4000/v1\",\n envKey: \"LITELLM_API_KEY\",\n website: \"https://litellm.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (self-hosted)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"browserai\",\n name: \"BrowserAI\",\n baseURL: \"https://api.browserai.com/v1\",\n envKey: \"BROWSERAI_API_KEY\",\n website: \"https://browserai.com\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"llmgateway\",\n name: \"LLM Gateway\",\n baseURL: \"https://api.llmgateway.io/v1\",\n envKey: \"LLMGATEWAY_API_KEY\",\n website: \"https://llmgateway.io\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"adaline\",\n name: \"Adaline AI Gateway\",\n baseURL: \"https://gateway.adaline.ai/v1\",\n envKey: \"ADALINE_API_KEY\",\n website: \"https://adaline.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"routeway\",\n name: \"Routeway\",\n baseURL: \"https://api.routeway.ai/v1\",\n envKey: \"ROUTEWAY_API_KEY\",\n website: \"https://routeway.ai\",\n apiType: \"openai-compatible\",\n freeTier: \":free zero-price models\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"anyscale\",\n name: \"Anyscale Endpoints\",\n baseURL: \"https://api.endpoints.anyscale.com/v1\",\n envKey: \"ANYSCALE_API_KEY\",\n website: \"https://anyscale.com\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"banana\",\n name: \"Banana ML\",\n baseURL: \"https://api.banana.dev/v1\",\n envKey: \"BANANA_API_KEY\",\n website: \"https://banana.dev\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"cuttle\",\n name: \"Cuttle AI\",\n baseURL: \"https://api.cuttle.ai/v1\",\n envKey: \"CUTTLE_API_KEY\",\n website: \"https://cuttle.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"infermatic\",\n name: \"Infermatic AI\",\n baseURL: \"https://api.infermatic.ai/v1\",\n envKey: \"INFERMATIC_API_KEY\",\n website: \"https://infermatic.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"mancer\",\n name: \"Mancer AI\",\n baseURL: \"https://api.mancer.ai/v1\",\n envKey: \"MANCER_API_KEY\",\n website: \"https://mancer.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"monsterapi\",\n name: \"Monster API\",\n baseURL: \"https://llm.monsterapi.ai/v1\",\n envKey: \"MONSTER_API_KEY\",\n website: \"https://monsterapi.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"dreamfactory\",\n name: \"DreamFactory AI\",\n baseURL: \"https://api.dreamfactory.ai/v1\",\n envKey: \"DREAMFACTORY_API_KEY\",\n website: \"https://dreamfactory.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"anya\",\n name: \"Anya AI\",\n baseURL: \"https://api.anya.ai/v1\",\n envKey: \"ANYA_API_KEY\",\n website: \"https://anya.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"khoj\",\n name: \"Khoj AI\",\n baseURL: \"https://api.khoj.dev/v1\",\n envKey: \"KHOJ_API_KEY\",\n website: \"https://khoj.dev\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"ollama\",\n name: \"Ollama\",\n baseURL: \"http://localhost:11434/v1\",\n website: \"https://ollama.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (self-hosted), 100+ models\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"ollama-cloud\",\n name: \"Ollama Cloud\",\n baseURL: \"https://api.ollama.com\",\n envKey: \"OLLAMA_API_KEY\",\n website: \"https://ollama.com/pricing\",\n apiType: \"openai-compatible\",\n freeTier: \"Free cloud usage\",\n permanentFree: true,\n },\n {\n id: \"lm-studio\",\n name: \"LM Studio\",\n baseURL: \"http://localhost:1234/v1\",\n website: \"https://lmstudio.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"llamacpp\",\n name: \"llama.cpp\",\n baseURL: \"http://localhost:8080/v1\",\n website: \"https://github.com/ggml-org/llama.cpp\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"vllm\",\n name: \"vLLM\",\n baseURL: \"http://localhost:8000/v1\",\n website: \"https://github.com/vllm-project/vllm\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (self-hosted)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"localai\",\n name: \"LocalAI\",\n baseURL: \"http://localhost:8080/v1\",\n website: \"https://localai.io\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (self-hosted)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"text-gen-ui\",\n name: \"Text Generation WebUI (oobabooga)\",\n baseURL: \"http://localhost:5000/v1\",\n website: \"https://github.com/oobabooga/text-generation-webui\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"tabbyapi\",\n name: \"TabbyAPI\",\n baseURL: \"http://localhost:5000/v1\",\n website: \"https://github.com/theroyallab/tabbyAPI\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"aphrodite\",\n name: \"Aphrodite Engine\",\n baseURL: \"http://localhost:2242/v1\",\n website: \"https://github.com/PygmalionAI/aphrodite-engine\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"kobold\",\n name: \"KoboldCpp\",\n baseURL: \"http://localhost:5001/v1\",\n website: \"https://github.com/LostRuins/koboldcpp\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"gpt4all\",\n name: \"GPT4All\",\n baseURL: \"http://localhost:4891/v1\",\n website: \"https://gpt4all.io\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local), CPU-only\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"jan\",\n name: \"Jan.ai\",\n baseURL: \"http://localhost:1337/v1\",\n website: \"https://jan.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local), offline\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"atomic-chat\",\n name: \"Atomic Chat\",\n baseURL: \"http://127.0.0.1:1337/v1\",\n website: \"https://atomic.chat\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"aleph-alpha\",\n name: \"Aleph Alpha\",\n baseURL: \"https://api.aleph-alpha.com/v1\",\n envKey: \"ALEPH_ALPHA_API_KEY\",\n website: \"https://www.aleph-alpha.com\",\n apiType: \"openai-compatible\",\n description: \"European sovereign AI\",\n },\n {\n id: \"deepset\",\n name: \"Deepset Cloud\",\n baseURL: \"https://api.deepset.ai/v1\",\n envKey: \"DEEPSET_API_KEY\",\n website: \"https://www.deepset.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"jina\",\n name: \"Jina AI\",\n baseURL: \"https://api.jina.ai/v1\",\n envKey: \"JINA_API_KEY\",\n website: \"https://jina.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"nomic\",\n name: \"Nomic AI\",\n baseURL: \"https://api.nomic.ai/v1\",\n envKey: \"NOMIC_API_KEY\",\n website: \"https://nomic.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"nlpcloud\",\n name: \"NLP Cloud\",\n baseURL: \"https://api.nlpcloud.io/v1\",\n envKey: \"NLPCLOUD_API_KEY\",\n website: \"https://nlpcloud.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Trial credits available\",\n },\n {\n id: \"sakana\",\n name: \"Sakana AI\",\n baseURL: \"https://api.sakana.ai/v1\",\n envKey: \"SAKANA_API_KEY\",\n website: \"https://sakana.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"stockmark\",\n name: \"Stockmark\",\n baseURL: \"https://api.stockmark.ai/v1\",\n envKey: \"STOCKMARK_API_KEY\",\n website: \"https://stockmark.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"elyza\",\n name: \"Elyza LLM\",\n baseURL: \"https://api.elyza.ai/v1\",\n envKey: \"ELYZA_API_KEY\",\n website: \"https://elyza.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"naver\",\n name: \"Naver HyperCLOVA\",\n baseURL: \"https://clovastudio.apigw.ntruss.com/v1\",\n envKey: \"NAVER_API_KEY\",\n website: \"https://www.ncloud.com\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"kakao\",\n name: \"Kakao Brain\",\n baseURL: \"https://api.kakaobrain.com/v1\",\n envKey: \"KAKAO_API_KEY\",\n website: \"https://kakaobrain.com\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"lg-ai\",\n name: \"LG AI (EXAONE)\",\n baseURL: \"https://api.lg-ai.com/v1\",\n envKey: \"LG_AI_API_KEY\",\n website: \"https://lg-ai.com\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"aispace\",\n name: \"AI Singapore (SEA-LION)\",\n baseURL: \"https://api.aisingapore.org/v1\",\n envKey: \"AISG_API_KEY\",\n website: \"https://aisingapore.org\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"core42\",\n name: \"Core42 / G42 (UAE)\",\n baseURL: \"https://api.core42.ai/v1\",\n envKey: \"CORE42_API_KEY\",\n website: \"https://core42.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"tii\",\n name: \"TII (Falcon / UAE)\",\n baseURL: \"https://api.tii.ae/v1\",\n envKey: \"TII_API_KEY\",\n website: \"https://tii.ae\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"froog\",\n name: \"Froog AI\",\n baseURL: \"https://api.froog.ai/v1\",\n envKey: \"FROOG_API_KEY\",\n website: \"https://froog.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"mendable\",\n name: \"Mendable\",\n baseURL: \"https://api.mendable.ai/v1\",\n envKey: \"MENDABLE_API_KEY\",\n website: \"https://mendable.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"fiddle\",\n name: \"Fiddle AI\",\n baseURL: \"https://api.fiddle.ai/v1\",\n envKey: \"FIDDLE_API_KEY\",\n website: \"https://fiddle.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"triton\",\n name: \"Triton AI\",\n baseURL: \"https://api.triton.ai/v1\",\n envKey: \"TRITON_API_KEY\",\n website: \"https://triton.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"shuttleai\",\n name: \"Shuttle AI\",\n baseURL: \"https://api.shuttle.ai/v1\",\n envKey: \"SHUTTLE_API_KEY\",\n website: \"https://shuttle.ai\",\n apiType: \"openai-compatible\",\n },\n {\n id: \"chatllm\",\n name: \"ChatLLM (Abacus)\",\n baseURL: \"https://api.chatllm.com/v1\",\n envKey: \"CHATLLM_API_KEY\",\n website: \"https://chatllm.com\",\n apiType: \"openai-compatible\",\n },\n\n // ── From provider-search.md: US/European Providers ──────────────\n {\n id: \"kimchi\",\n name: \"Kimchi Inference\",\n baseURL: \"https://llm.kimchi.dev/openai/v1\",\n envKey: \"KIMCHI_API_KEY\",\n website: \"https://kimchi.dev\",\n apiType: \"openai-compatible\",\n freeTier: \"$50/mo free credits for developers\",\n permanentFree: true,\n },\n {\n id: \"tokenrouter\",\n name: \"TokenRouter\",\n baseURL: \"https://api.tokenrouter.io/v1\",\n envKey: \"TOKENROUTER_API_KEY\",\n website: \"https://tokenrouter.io\",\n apiType: \"openai-compatible\",\n freeTier: \"Free MiniMax M3 API access\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"aerolink\",\n name: \"Aerolink\",\n baseURL: \"https://aerolink.lat\",\n envKey: \"AEROLINK_API_KEY\",\n website: \"https://aerolink.lat\",\n apiType: \"openai-compatible\",\n freeTier: \"Free $350+ rolling credits ($10/5 hours)\",\n permanentFree: true,\n },\n {\n id: \"b-ai\",\n name: \"B.AI\",\n baseURL: \"https://api.b.ai/v1\",\n envKey: \"BAI_API_KEY\",\n website: \"https://b.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"500K credits on signup, no card\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"lightning-ai\",\n name: \"Lightning AI Models\",\n baseURL: \"https://api.lightning.ai/v1\",\n envKey: \"LIGHTNING_API_KEY\",\n website: \"https://lightning.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"15 free credits/mo ($15 value, up to 30M tokens)\",\n permanentFree: true,\n },\n {\n id: \"aion-labs\",\n name: \"Aion Labs\",\n baseURL: \"https://api.aionlabs.ai/v1\",\n envKey: \"AION_API_KEY\",\n website: \"https://aionlabs.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Daily free token allowance on registration\",\n permanentFree: true,\n },\n {\n id: \"opencode-zen\",\n name: \"OpenCode Zen\",\n baseURL: \"https://opencode.ai/zen/v1\",\n envKey: \"OPENCODE_ZEN_API_KEY\",\n website: \"https://opencode.ai/zen\",\n apiType: \"openai-compatible\",\n freeTier: \"Free curated gateway\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"lepton\",\n name: \"Lepton AI\",\n baseURL: \"https://api.lepton.ai/v1\",\n envKey: \"LEPTON_API_KEY\",\n website: \"https://lepton.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"$10 free trial credits for developers\",\n permanentFree: true,\n },\n {\n id: \"friendli\",\n name: \"Friendli AI\",\n baseURL: \"https://api.friendli.ai/v1\",\n envKey: \"FRIENDLI_API_KEY\",\n website: \"https://friendli.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"$10 free trial credits\",\n permanentFree: true,\n },\n {\n id: \"cerebrium\",\n name: \"Cerebrium\",\n baseURL: \"https://api.cerebrium.ai/v1\",\n envKey: \"CEREBRIUM_API_KEY\",\n website: \"https://cerebrium.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"$30 free trial credits\",\n permanentFree: true,\n },\n {\n id: \"requesty\",\n name: \"Requesty\",\n baseURL: \"https://api.requesty.ai/v1\",\n envKey: \"REQUESTY_API_KEY\",\n website: \"https://requesty.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Free monthly credit allocations\",\n permanentFree: true,\n },\n {\n id: \"zenmux\",\n name: \"ZenMux\",\n baseURL: \"https://api.zenmux.com/v1\",\n envKey: \"ZENMUX_API_KEY\",\n website: \"https://zenmux.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free sandbox API tier\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"crofai\",\n name: \"CrofAI\",\n baseURL: \"https://api.crofai.com/v1\",\n envKey: \"CROFAI_API_KEY\",\n website: \"https://crofai.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Promo credits at registration\",\n permanentFree: true,\n },\n {\n id: \"fastrouter\",\n name: \"FastRouter\",\n baseURL: \"https://api.fastrouter.ai/v1\",\n envKey: \"FASTROUTER_API_KEY\",\n website: \"https://fastrouter.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Trial credits\",\n permanentFree: true,\n },\n {\n id: \"aiqu\",\n name: \"AiQu\",\n baseURL: \"https://api.aiqu.se/v1\",\n envKey: \"AIQU_API_KEY\",\n website: \"https://aiqu.se\",\n apiType: \"openai-compatible\",\n freeTier: \"Swedish cloud free sandbox\",\n noCc: true,\n permanentFree: true,\n description: \"EU sovereign hosted\",\n },\n {\n id: \"airon\",\n name: \"Airon\",\n baseURL: \"https://api.airon.io/v1\",\n envKey: \"AIRON_API_KEY\",\n website: \"https://airon.io\",\n apiType: \"openai-compatible\",\n freeTier: \"Serverless trial credits\",\n permanentFree: true,\n },\n {\n id: \"aki\",\n name: \"AKI.IO\",\n baseURL: \"https://api.aki.io/v1\",\n envKey: \"AKI_API_KEY\",\n website: \"https://aki.io\",\n apiType: \"openai-compatible\",\n freeTier: \"10 EUR trial credits\",\n permanentFree: true,\n },\n {\n id: \"ark-labs\",\n name: \"ARK Labs\",\n baseURL: \"https://api.arklabs.ai/v1\",\n envKey: \"ARK_API_KEY\",\n website: \"https://arklabs.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"EU sovereign free tier\",\n noCc: true,\n permanentFree: true,\n description: \"Regulated environment models\",\n },\n {\n id: \"beam\",\n name: \"Beam\",\n baseURL: \"https://api.beam.cloud/v1\",\n envKey: \"BEAM_API_KEY\",\n website: \"https://beam.cloud\",\n apiType: \"openai-compatible\",\n freeTier: \"Serverless GPU free trial tier\",\n permanentFree: true,\n },\n {\n id: \"bentocloud\",\n name: \"BentoCloud\",\n baseURL: \"https://api.bentocloud.com/v1\",\n envKey: \"BENTOML_API_KEY\",\n website: \"https://bentoml.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free trial credits\",\n permanentFree: true,\n },\n {\n id: \"berget\",\n name: \"Berget AI\",\n baseURL: \"https://api.berget.ai/v1\",\n envKey: \"BERGET_API_KEY\",\n website: \"https://berget.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"EU sovereign free tier\",\n noCc: true,\n permanentFree: true,\n description: \"EU compliant models\",\n },\n {\n id: \"cortecs\",\n name: \"Cortecs AI\",\n baseURL: \"https://api.cortecs.ai/v1\",\n envKey: \"CORTECS_API_KEY\",\n website: \"https://cortecs.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Gateway free trial\",\n permanentFree: true,\n },\n {\n id: \"eurouter\",\n name: \"EUrouter\",\n baseURL: \"https://api.eurouter.com/v1\",\n envKey: \"EUROUTER_API_KEY\",\n website: \"https://eurouter.com\",\n apiType: \"openai-compatible\",\n freeTier: \"1K requests/month free\",\n noCc: true,\n permanentFree: true,\n description: \"100+ GDPR models\",\n },\n {\n id: \"infercom\",\n name: \"Infercom\",\n baseURL: \"https://api.infercom.ai/v1\",\n envKey: \"INFERCOM_API_KEY\",\n website: \"https://infercom.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Free tier, EU hosted\",\n noCc: true,\n permanentFree: true,\n description: \"Sovereign open models\",\n },\n {\n id: \"ionrouter\",\n name: \"IonRouter\",\n baseURL: \"https://api.ionrouter.com/v1\",\n envKey: \"IONROUTER_API_KEY\",\n website: \"https://ionrouter.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free tier options\",\n permanentFree: true,\n },\n {\n id: \"lambda\",\n name: \"Lambda AI\",\n baseURL: \"https://api.lambdalabs.com/v1\",\n envKey: \"LAMBDA_API_KEY\",\n website: \"https://lambdalabs.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Trial cloud credits\",\n permanentFree: true,\n },\n {\n id: \"octoai\",\n name: \"OctoAI\",\n baseURL: \"https://api.octoai.cloud/v1\",\n envKey: \"OCTOAI_API_KEY\",\n website: \"https://octoai.cloud\",\n apiType: \"openai-compatible\",\n freeTier: \"Free trial credits\",\n permanentFree: true,\n },\n {\n id: \"prem\",\n name: \"Prem AI\",\n baseURL: \"https://api.premai.io/v1\",\n envKey: \"PREM_API_KEY\",\n website: \"https://premai.io\",\n apiType: \"openai-compatible\",\n freeTier: \"Free tier\",\n permanentFree: true,\n },\n {\n id: \"synexa\",\n name: \"Synexa\",\n baseURL: \"https://api.synexa.io/v1\",\n envKey: \"SYNEXA_API_KEY\",\n website: \"https://synexa.io\",\n apiType: \"openai-compatible\",\n freeTier: \"Trial credits, $0.0015/image\",\n permanentFree: true,\n },\n {\n id: \"tokenware\",\n name: \"Tokenware\",\n baseURL: \"https://api.tokenware.io/v1\",\n envKey: \"TOKENWARE_API_KEY\",\n website: \"https://tokenware.io\",\n apiType: \"openai-compatible\",\n freeTier: \"Free trial credits, 200+ models\",\n permanentFree: true,\n },\n {\n id: \"voyage\",\n name: \"Voyage AI\",\n baseURL: \"https://api.voyageai.com/v1\",\n envKey: \"VOYAGE_API_KEY\",\n website: \"https://voyageai.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free tier credits available\",\n permanentFree: true,\n description: \"High quality embeddings\",\n },\n {\n id: \"featherless\",\n name: \"Featherless AI\",\n baseURL: \"https://api.featherless.ai/v1\",\n envKey: \"FEATHERLESS_API_KEY\",\n website: \"https://featherless.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Free tier trial, 200+ open-source models\",\n permanentFree: true,\n },\n {\n id: \"morph\",\n name: \"Morph\",\n baseURL: \"https://api.morph.chat/v1\",\n envKey: \"MORPH_API_KEY\",\n website: \"https://morph.chat\",\n apiType: \"openai-compatible\",\n freeTier: \"Agentic free credits\",\n permanentFree: true,\n },\n {\n id: \"predibase\",\n name: \"Predibase\",\n baseURL: \"https://connect.predibase.com/v1\",\n envKey: \"PREDIBASE_API_KEY\",\n website: \"https://predibase.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Free credits on signup\",\n permanentFree: true,\n description: \"Custom fine-tuned LoRA models\",\n },\n {\n id: \"ferryapi\",\n name: \"FerryAPI\",\n baseURL: \"https://api.ferryapi.com/v1\",\n envKey: \"FERRYAPI_API_KEY\",\n website: \"https://ferryapi.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Prepay signup bonus\",\n permanentFree: true,\n },\n {\n id: \"fastpivot\",\n name: \"Fast Pivot\",\n baseURL: \"https://api.fastpivot.io/v1\",\n envKey: \"FASTPIVOT_API_KEY\",\n website: \"https://fastpivot.io\",\n apiType: \"openai-compatible\",\n freeTier: \"Prepaid credit trial\",\n permanentFree: true,\n },\n {\n id: \"general-compute\",\n name: \"General Compute\",\n baseURL: \"https://api.generalcompute.com/v1\",\n envKey: \"GENERAL_COMPUTE_API_KEY\",\n website: \"https://generalcompute.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Trial compute hours\",\n permanentFree: true,\n description: \"Latency-critical coding models\",\n },\n {\n id: \"coreweave\",\n name: \"CoreWeave\",\n baseURL: \"https://api.coreweave.com/v1\",\n envKey: \"COREWEAVE_API_KEY\",\n website: \"https://coreweave.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Trial cloud credits\",\n permanentFree: true,\n },\n {\n id: \"venice\",\n name: \"Venice.ai\",\n baseURL: \"https://api.venice.ai/api/v1\",\n envKey: \"VENICE_API_KEY\",\n website: \"https://venice.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Daily limits allowed for free tier\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"packet\",\n name: \"Packet.ai\",\n baseURL: \"https://api.packet.ai/v1\",\n envKey: \"PACKET_API_KEY\",\n website: \"https://packet.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Billed per token, trial credits\",\n permanentFree: true,\n },\n {\n id: \"verda\",\n name: \"Verda\",\n baseURL: \"https://api.verda.ai/v1\",\n envKey: \"VERDA_API_KEY\",\n website: \"https://verda.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Serverless trial\",\n permanentFree: true,\n description: \"European optimized LLMs\",\n },\n\n // ── From provider-search.md: Chinese/Asian Providers ────────────\n {\n id: \"yi\",\n name: \"Yi AI / 01.AI (零一万物)\",\n baseURL: \"https://api.01.ai/v1\",\n envKey: \"YI_API_KEY\",\n website: \"https://01.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Initial trial credits on signup\",\n permanentFree: true,\n },\n {\n id: \"mimo\",\n name: \"Xiaomi Mimo (小米)\",\n baseURL: \"https://api.mimo.xiaomi.com/v1\",\n envKey: \"MIMO_API_KEY\",\n website: \"https://mi.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Highly competitive pricing ($0.14/1M)\",\n permanentFree: true,\n },\n {\n id: \"streamlake\",\n name: \"StreamLake / Kuaishou (快手)\",\n baseURL: \"https://api.streamlake.com/v1\",\n envKey: \"STREAMLAKE_API_KEY\",\n website: \"https://streamlake.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Developer trial credits\",\n permanentFree: true,\n },\n {\n id: \"gitee-ai\",\n name: \"Gitee AI (码云)\",\n baseURL: \"https://ai.gitee.com/v1\",\n envKey: \"GITEE_AI_TOKEN\",\n website: \"https://gitee.com/ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Developer free sandbox tokens\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"linkos\",\n name: \"Linkos AI\",\n baseURL: \"https://api.linkos.ai/v1\",\n envKey: \"LINKOS_API_KEY\",\n website: \"https://linkos.ai\",\n apiType: \"openai-compatible\",\n freeTier: \"Gateway free trial\",\n permanentFree: true,\n },\n {\n id: \"baobaogou\",\n name: \"Baobaogou AI (抱抱狗)\",\n baseURL: \"https://api.baobaogou.com/v1\",\n envKey: \"BAOBAOGOU_API_KEY\",\n website: \"https://baobaogou.com\",\n apiType: \"openai-compatible\",\n freeTier: \"Regional trial credits\",\n permanentFree: true,\n },\n\n // ── From provider-search.md: Local / Self-Hosted ───────────────\n {\n id: \"aegisgate\",\n name: \"AegisGate\",\n baseURL: \"http://localhost:18080/v1\",\n website: \"https://github.com/ax128/AegisGate\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (self-hosted security gateway)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"freellmapi\",\n name: \"FreeLLMAPI\",\n baseURL: \"http://localhost:3000/v1\",\n website: \"https://github.com/open-free-llm-api/freellm-hub-data\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (self-hosted proxy aggregator)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"sglang\",\n name: \"SGLang Server\",\n baseURL: \"http://localhost:30000/v1\",\n website: \"https://github.com/sgl-project/sglang\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (self-hosted), latency-optimized\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"tabby\",\n name: \"Tabby AI\",\n baseURL: \"http://localhost:8080/v1\",\n website: \"https://github.com/TabbyML/tabby\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (self-hosted code completion)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"mindspore\",\n name: \"MindSpore\",\n baseURL: \"http://localhost:8080/v1\",\n website: \"https://github.com/mindspore-ai/mindspore\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local deep learning framework)\",\n noCc: true,\n permanentFree: true,\n },\n {\n id: \"onnxruntime\",\n name: \"ONNX Runtime\",\n baseURL: \"http://localhost:8090/v1\",\n website: \"https://github.com/microsoft/onnxruntime\",\n apiType: \"openai-compatible\",\n freeTier: \"Unlimited (local), multi-platform hardware acceleration\",\n noCc: true,\n permanentFree: true,\n },\n]\n\nexport function getProviders(): Provider[] {\n return PROVIDERS\n}\n\nexport function getProvider(id: string): Provider | undefined {\n return PROVIDERS.find(p => p.id === id)\n}\n\nexport function searchProviders(query: string): Provider[] {\n const q = query.toLowerCase()\n return PROVIDERS.filter(p =>\n p.id.toLowerCase().includes(q) ||\n p.name.toLowerCase().includes(q)\n )\n}\n\nexport interface CuratedFavorite {\n id: string\n tags: string[]\n description: string\n}\n\nexport const CURATED_FAVORITES: CuratedFavorite[] = [\n { id: \"groq\", tags: [\"coding\", \"fast\", \"free\", \"no-cc\"], description: \"Fastest free inference, great for code\" },\n { id: \"cerebras\", tags: [\"coding\", \"fast\", \"free\", \"no-cc\"], description: \"World's fastest inference, free tier\" },\n { id: \"google-gemini\", tags: [\"coding\", \"free\", \"vision\", \"reasoning\", \"no-cc\"], description: \"Gemini 2.5 Flash/Pro, free generous tier\" },\n { id: \"github-models\", tags: [\"coding\", \"free\", \"no-cc\"], description: \"Free GPT-4o, Claude, Gemini, DeepSeek\" },\n { id: \"deepseek\", tags: [\"coding\", \"reasoning\", \"free\", \"no-cc\"], description: \"Best open coding model (R1, V3)\" },\n { id: \"mistral\", tags: [\"coding\", \"free\", \"reasoning\"], description: \"Codestral + Mistral models, free tier\" },\n { id: \"cohere\", tags: [\"coding\", \"free\", \"reasoning\", \"no-cc\"], description: \"Command R7, free 1M tok/month\" },\n { id: \"together\", tags: [\"coding\", \"free\", \"fast\", \"reasoning\", \"no-cc\"], description: \"100+ open models, generous free tier\" },\n { id: \"openrouter\", tags: [\"coding\", \"free\", \"no-cc\", \"reasoning\", \"vision\"], description: \"Gateway to 200+ models, free tier\" },\n { id: \"sambanova\", tags: [\"coding\", \"free\", \"fast\", \"no-cc\"], description: \"Fast free inference, good for code\" },\n { id: \"fireworks\", tags: [\"coding\", \"free\", \"fast\", \"no-cc\"], description: \"Fast free inference, open models\" },\n { id: \"deepinfra\", tags: [\"coding\", \"free\", \"fast\"], description: \"Fast free inference, good selection\" },\n { id: \"hyperbolic\", tags: [\"coding\", \"free\", \"reasoning\"], description: \"Free tier with reasoning models\" },\n { id: \"novita\", tags: [\"coding\", \"free\"], description: \"Free tier, 30+ models available\" },\n { id: \"nebius\", tags: [\"coding\", \"free\", \"fast\"], description: \"Free tier, fast inference\" },\n { id: \"perplexity\", tags: [\"coding\", \"free\", \"reasoning\"], description: \"Sonar models, free 5 RPD\" },\n { id: \"lepton\", tags: [\"coding\", \"free\"], description: \"Free tier, good model selection\" },\n { id: \"ai21\", tags: [\"coding\", \"free\"], description: \"Jamba 1.6, free 200 RPM\" },\n { id: \"arcee\", tags: [\"coding\", \"free\", \"no-cc\"], description: \"Free, no credit card needed\" },\n { id: \"huggingface\", tags: [\"coding\", \"free\", \"no-cc\", \"reasoning\", \"vision\"], description: \"100K+ free models via Inference API\" },\n { id: \"opencode-zen\", tags: [\"coding\", \"free\", \"fast\", \"no-cc\"], description: \"OpenCode's built-in free proxy\" },\n { id: \"grok-xai\", tags: [\"coding\", \"free\", \"reasoning\"], description: \"Grok-2/3 free with X Premium\" },\n { id: \"replicate\", tags: [\"coding\", \"free\", \"vision\"], description: \"Free trial, 100K+ community models\" },\n { id: \"anyscale\", tags: [\"coding\", \"free\", \"fast\"], description: \"Serverless endpoint, free credits\" },\n { id: \"anthropic\", tags: [\"coding\", \"reasoning\"], description: \"Claude 4 Sonnet/Opus, best-in-class coding\" },\n]\n\nexport function getRecommendedProviderIds(): string[] {\n return CURATED_FAVORITES.map(f => f.id)\n}\n\nexport function getRecommendedProviders(): Provider[] {\n const ids = new Set(getRecommendedProviderIds())\n return PROVIDERS.filter(p => ids.has(p.id))\n}\n\nexport function getFavoritesByTag(tag: string): Provider[] {\n const ids = new Set(\n CURATED_FAVORITES.filter(f => f.tags.includes(tag)).map(f => f.id),\n )\n return PROVIDERS.filter(p => ids.has(p.id))\n}\n\nexport function getProviderByEnvKey(envKey: string): Provider | undefined {\n return PROVIDERS.find(p => p.envKey === envKey)\n}\n\nexport function getFreeProviders(): Provider[] {\n return PROVIDERS.filter(p => p.permanentFree && p.noCc)\n}\n\nexport function getProvidersSortedByCost(): Provider[] {\n return [...PROVIDERS].sort((a, b) => {\n const aScore = a.permanentFree && a.noCc ? 0 : a.permanentFree ? 1 : a.freeTier ? 2 : 3\n const bScore = b.permanentFree && b.noCc ? 0 : b.permanentFree ? 1 : b.freeTier ? 2 : 3\n return aScore - bScore\n })\n}"}}}} |