[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-74336":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":23,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":40,"readmeContent":41,"aiSummary":42,"trendingCount":16,"starSnapshotCount":16,"syncStatus":43,"lastSyncTime":44,"discoverSource":45},74336,"awesome-free-llm-apis","mnfst\u002Fawesome-free-llm-apis","mnfst","List of Permanent Free LLM API  (API Keys)","",null,"JavaScript",4951,470,66,5,0,86,271,703,258,110.02,"Creative Commons Zero v1.0 Universal",false,"main",true,[27,28,29,30,31,32,33,34,35,36,37,38,39],"ai-agents","anthropic","awesome","awesome-list","gemini","llm","llm-router","llm-routing","ollama","openai","openclaw","openclaw-plugin","router","2026-06-12 04:01:14","\u003Ch1 align=\"center\">\n\t\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fmnfst\u002Fawesome-free-llm-apis\">\n\t\t\u003Cimg src=\"media\u002Fawesome-free-llm-apis.png\" width=\"500\" alt=\"Awesome Free LLM APIs\">\n\t\u003C\u002Fa>\n\u003C\u002Fh1>\n\n\u003Cp align=\"center\">\n\t\u003Ca href=\"https:\u002F\u002Fawesome.re\">\n\t\t\u003Cimg src=\"https:\u002F\u002Fawesome.re\u002Fbadge-flat2.svg\" alt=\"Awesome\">\n\t\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">LLM APIs with permanent free tiers for text inference.\u003C\u002Fp>\n\n\u003Cp align=\"center\">\u003Csub>All endpoints are OpenAI SDK-compatible unless noted. Each link points to the provider's API key page.\u003C\u002Fsub>\u003C\u002Fp>\n\n## Contents\n\n- [Provider APIs](#provider-apis)\n- [Inference providers](#inference-providers)\n- [Glossary](#glossary)\n\n## Provider APIs\n\nAPIs run by the companies that train or fine-tune the models themselves.\n\n### [Cohere](https:\u002F\u002Fdashboard.cohere.com\u002Fapi-keys) 🇨🇦\n\nFree \"Trial\" API key, no credit card. 1,000 API calls\u002Fmonth. Non-commercial use only.\n\nBase URL: `https:\u002F\u002Fapi.cohere.com\u002Fv2`\n\n| Model Name       | Context | Max Output | Modality                  | Rate Limit       |\n| ---------------- | ------- | ---------- | ------------------------- | ---------------- |\n| Command A (111B) | 256K    | 4K         | Text                      | 20 RPM           |\n| Command R+       | 128K    | 4K         | Text                      | 20 RPM           |\n| Command R        | 128K    | 4K         | Text                      | 20 RPM           |\n| Command R7B      | 128K    | 4K         | Text                      | 20 RPM           |\n| Embed 4          | —       | —          | Embeddings (Text + Image) | 2,000 inputs\u002Fmin |\n| Rerank 3.5       | —       | —          | Reranking                 | 10 RPM           |\n\n### [Google Gemini](https:\u002F\u002Faistudio.google.com\u002Fapp\u002Fapikey) 🇺🇸\n\nFree tier unavailable in EU\u002FUK\u002FSwitzerland. Free-tier prompts may be used by Google to improve products. [^1]\n\nBase URL: `https:\u002F\u002Fgenerativelanguage.googleapis.com\u002Fv1beta`\n\n| Model Name            | Context | Max Output | Modality                     | Rate Limit        |\n| --------------------- | ------- | ---------- | ---------------------------- | ----------------- |\n| Gemini 2.5 Flash      | 1M      | 65K        | Text + Image + Audio + Video | 10 RPM, 250 RPD   |\n| Gemini 2.5 Flash-Lite | 1M      | 65K        | Text + Image + Audio + Video | 15 RPM, 1,000 RPD |\n\n### [Mistral AI](https:\u002F\u002Fconsole.mistral.ai\u002Fapi-keys) 🇫🇷\n\nFree \"Experiment\" plan, no credit card. ~1B tokens\u002Fmonth.\n\nBase URL: `https:\u002F\u002Fapi.mistral.ai\u002Fv1`\n\n| Model Name         | Context | Max Output | Modality            | Rate Limit       |\n| ------------------ | ------- | ---------- | ------------------- | ---------------- |\n| Mistral Small 4    | 256K    | 256K       | Text + Image + Code | ~1 RPS, 500K TPM |\n| Mistral Medium 3   | 128K    | 128K       | Text                | ~1 RPS, 500K TPM |\n| Mistral Large 3    | 256K    | 256K       | Text                | ~1 RPS, 500K TPM |\n| Mistral Nemo (12B) | 128K    | 128K       | Text                | ~1 RPS, 500K TPM |\n| Codestral          | 256K    | 256K       | Code                | ~1 RPS, 500K TPM |\n| Pixtral Large      | 128K    | 128K       | Text + Image        | ~1 RPS, 500K TPM |\n\n### [Z AI (Zhipu AI)](https:\u002F\u002Fopen.bigmodel.cn\u002Fusercenter\u002Fapikeys) 🇨🇳\n\nPermanent free models, no credit card required.\n\nBase URL: `https:\u002F\u002Fopen.bigmodel.cn\u002Fapi\u002Fpaas\u002Fv4`\n\n| Model Name     | Context | Max Output | Modality     | Rate Limit           |\n| -------------- | ------- | ---------- | ------------ | -------------------- |\n| GLM-4.7-Flash  | 200K    | 128K       | Text         | 1 concurrent request |\n| GLM-4.5-Flash  | 128K    | ~8K        | Text         | 1 concurrent request |\n| GLM-4.6V-Flash | 128K    | ~4K        | Text + Image | 1 concurrent request |\n\n## Inference providers\n\nThird-party platforms that host open-weight models from various sources.\n\n### [Cerebras](https:\u002F\u002Fcloud.cerebras.ai\u002F) 🇺🇸\n\nFree tier, no credit card. Ultra-fast inference (~2,600 tok\u002Fs). 1M tokens\u002Fday cap.\n\nBase URL: `https:\u002F\u002Fapi.cerebras.ai\u002Fv1`\n\n| Model Name                     | Context           | Max Output | Modality | Rate Limit                 |\n| ------------------------------ | ----------------- | ---------- | -------- | -------------------------- |\n| llama3.1-8b                    | 128K (8K on free) | 8K         | Text     | 30 RPM, 14,400 RPD, 1M TPD |\n| gpt-oss-120b                   | 128K (8K on free) | 8K         | Text     | 30 RPM, 14,400 RPD, 1M TPD |\n| qwen-3-235b-a22b-instruct-2507 | 131K (8K on free) | 8K         | Text     | 30 RPM, 14,400 RPD, 1M TPD |\n| zai-glm-4.7                    | 128K (8K on free) | 8K         | Text     | 10 RPM, 100 RPD, 1M TPD    |\n\n### [Cloudflare Workers AI](https:\u002F\u002Fdash.cloudflare.com\u002Fprofile\u002Fapi-tokens) 🇺🇸\n\n10,000 Neurons\u002Fday free. 50+ models available on free tier.\n\nBase URL: `https:\u002F\u002Fapi.cloudflare.com\u002Fclient\u002Fv4\u002Faccounts\u002F{account_id}\u002Fai\u002Frun`\n\n| Model Name                                   | Context   | Max Output        | Modality                       | Rate Limit               |\n| -------------------------------------------- | --------- | ----------------- | ------------------------------ | ------------------------ |\n| @cf\u002Fmeta\u002Fllama-3.3-70b-instruct-fp8-fast     | 131K      | Shared w\u002F context | Text                           | 10K neurons\u002Fday (shared) |\n| @cf\u002Fmeta\u002Fllama-3.1-8b-instruct-fp8-fast      | 131K      | Shared w\u002F context | Text                           | 10K neurons\u002Fday (shared) |\n| @cf\u002Fmeta\u002Fllama-3.2-11b-vision-instruct       | 131K      | Shared w\u002F context | Text + Vision                  | 10K neurons\u002Fday (shared) |\n| @cf\u002Fmeta\u002Fllama-4-scout-17b-16e-instruct      | Up to 10M | Shared w\u002F context | Multimodal                     | 10K neurons\u002Fday (shared) |\n| @cf\u002Fmistralai\u002Fmistral-small-3.1-24b-instruct | 128K      | Shared w\u002F context | Text                           | 10K neurons\u002Fday (shared) |\n| @cf\u002Fgoogle\u002Fgemma-4-26b-a4b-it                | 256K      | Shared w\u002F context | Text                           | 10K neurons\u002Fday (shared) |\n| @cf\u002Fqwen\u002Fqwq-32b                             | 32K       | Shared w\u002F context | Text                           | 10K neurons\u002Fday (shared) |\n| @cf\u002Fdeepseek-ai\u002Fdeepseek-r1-distill-qwen-32b | 32K       | Shared w\u002F context | Text                           | 10K neurons\u002Fday (shared) |\n| + 42 more models                             | Varies    | Varies            | Text, Image, Audio, Embeddings | 10K neurons\u002Fday (shared) |\n\n### [GitHub Models](https:\u002F\u002Fgithub.com\u002Fmarketplace\u002Fmodels) 🇺🇸\n\nFree prototyping for all GitHub users. 45+ models. Per-request limits (8K in \u002F 4K out).\n\nBase URL: `https:\u002F\u002Fmodels.inference.ai.azure.com`\n\n| Model Name                | Context | Max Output | Modality         | Rate Limit      |\n| ------------------------- | ------- | ---------- | ---------------- | --------------- |\n| gpt-4.1                   | 1M      | 32K        | Text             | 10 RPM, 50 RPD  |\n| gpt-4.1-mini              | 1M      | 32K        | Text             | 15 RPM, 150 RPD |\n| gpt-4o                    | 128K    | 16K        | Text + Vision    | 10 RPM, 50 RPD  |\n| o3-mini                   | 200K    | 100K       | Text (reasoning) | 10 RPM, 50 RPD  |\n| o4-mini                   | 200K    | 100K       | Text (reasoning) | 10 RPM, 50 RPD  |\n| Llama-4-Scout-17B-16E     | 512K    | ~4K        | Text + Vision    | 15 RPM, 150 RPD |\n| Llama-4-Maverick-17B-128E | 256K    | ~4K        | Text + Vision    | 10 RPM, 50 RPD  |\n| Meta-Llama-3.3-70B        | 131K    | ~4K        | Text             | 15 RPM, 150 RPD |\n| DeepSeek-R1               | 64K     | 8K         | Text (reasoning) | 15 RPM, 150 RPD |\n| Mistral-Small-3.1         | 128K    | ~4K        | Text + Vision    | 15 RPM, 150 RPD |\n| + 35 more models          | Varies  | Varies     | Text \u002F Image     | Varies by tier  |\n\n### [Groq](https:\u002F\u002Fconsole.groq.com\u002Fkeys) 🇺🇸\n\nFree tier, no credit card. Ultra-fast LPU inference. [^2]\n\nBase URL: `https:\u002F\u002Fapi.groq.com\u002Fopenai\u002Fv1`\n\n| Model Name                         | Context | Max Output | Modality      | Rate Limit         |\n| ---------------------------------- | ------- | ---------- | ------------- | ------------------ |\n| llama-3.3-70b-versatile            | 131K    | 32K        | Text          | 30 RPM, 14,400 RPD |\n| llama-3.1-8b-instant               | 131K    | 131K       | Text          | 30 RPM, 14,400 RPD |\n| llama-4-scout-17b-16e-instruct     | 131K    | 8K         | Text + Vision | 30 RPM, 14,400 RPD |\n| llama-4-maverick-17b-128e-instruct | 131K    | 8K         | Text + Vision | 15 RPM, 500 RPD    |\n| qwen3-32b                          | 131K    | 131K       | Text          | 30 RPM, 14,400 RPD |\n| gpt-oss-120b                       | 131K    | 32K        | Text          | 30 RPM, 14,400 RPD |\n| kimi-k2-instruct                   | 262K    | 262K       | Text          | 30 RPM, 14,400 RPD |\n| deepseek-r1-distill-70b            | 131K    | 8K         | Text          | 30 RPM, 14,400 RPD |\n| whisper-large-v3                   | —       | —          | Audio → Text  | 20 RPM, 2,000 RPD  |\n| whisper-large-v3-turbo             | —       | —          | Audio → Text  | 20 RPM, 2,000 RPD  |\n\n### [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fsettings\u002Ftokens) 🇺🇸\n\nFree Serverless Inference API + ~$0.10\u002Fmonth free credits. Thousands of models.\n\nBase URL: `https:\u002F\u002Fapi-inference.huggingface.co\u002Fmodels`\n\n| Model Name                      | Context | Max Output | Modality                       | Rate Limit                |\n| ------------------------------- | ------- | ---------- | ------------------------------ | ------------------------- |\n| Meta-Llama-3.1-8B-Instruct      | 128K    | ~4K        | Text                           | ~1,000 RPD                |\n| Mistral-7B-Instruct-v0.3        | 32K     | ~4K        | Text                           | ~1,000 RPD                |\n| Mixtral-8x7B-Instruct-v0.1      | 32K     | ~4K        | Text                           | ~1,000 RPD                |\n| Phi-3.5-mini-instruct           | 128K    | ~4K        | Text                           | ~1,000 RPD                |\n| Qwen2.5-7B-Instruct             | 131K    | ~4K        | Text                           | ~1,000 RPD                |\n| + thousands of community models | Varies  | Varies     | Text, Image, Audio, Embeddings | ~$0.10\u002Fmonth free credits |\n\n### [Kilo Code](https:\u002F\u002Fkilo.ai) 🇺🇸\n\nFree models with no credit card required. `kilo-auto\u002Ffree` auto-router routes to minimax\u002Fminimax-m2.5:free (80%) and stepfun\u002Fstep-3.5-flash:free (20%). [^5]\n\nBase URL: `https:\u002F\u002Fapi.kilo.ai\u002Fapi\u002Fgateway`\n\n| Model Name                             | Context | Max Output | Modality         | Rate Limit  |\n| -------------------------------------- | ------- | ---------- | ---------------- | ----------- |\n| bytedance-seed\u002Fdola-seed-2.0-pro:free  | —       | —          | Text             | ~200 req\u002Fhr |\n| x-ai\u002Fgrok-code-fast-1:optimized:free   | —       | —          | Text (code)      | ~200 req\u002Fhr |\n| nvidia\u002Fnemotron-3-super-120b-a12b:free | 262K    | 32K        | Text             | ~200 req\u002Fhr |\n| arcee-ai\u002Ftrinity-large-thinking:free   | —       | —          | Text (reasoning) | ~200 req\u002Fhr |\n| openrouter\u002Ffree                        | Varies  | Varies     | Text             | ~200 req\u002Fhr |\n\n### [LLM7.io](https:\u002F\u002Ftoken.llm7.io) 🇬🇧\n\nZero-friction API gateway. No registration needed for basic access. 30+ models.\n\nBase URL: `https:\u002F\u002Fapi.llm7.io\u002Fv1`\n\n| Model Name            | Context | Max Output | Modality         | Rate Limit              |\n| --------------------- | ------- | ---------- | ---------------- | ----------------------- |\n| deepseek-r1-0528      | —       | —          | Text (reasoning) | 30 RPM (120 with token) |\n| deepseek-v3-0324      | —       | —          | Text             | 30 RPM (120 with token) |\n| gemini-2.5-flash-lite | —       | —          | Text + Vision    | 30 RPM (120 with token) |\n| gpt-4o-mini           | —       | —          | Text + Vision    | 30 RPM (120 with token) |\n| mistral-small-3.1-24b | 32K     | —          | Text             | 30 RPM (120 with token) |\n| qwen2.5-coder-32b     | —       | —          | Text (code)      | 30 RPM (120 with token) |\n| + ~24 more models     | Varies  | Varies     | Text             | 30 RPM (120 with token) |\n\n### [ModelScope](https:\u002F\u002Fmodelscope.cn\u002Fmy\u002Fmyaccesstoken) 🇨🇳\n\nFree API-Inference for registered users. Requires Alibaba Cloud account binding + real-name verification. [^6]\n\nBase URL: `https:\u002F\u002Fapi-inference.modelscope.cn\u002Fv1`\n\n| Model Name                     | Context | Max Output | Modality         | Rate Limit                                 |\n| ------------------------------ | ------- | ---------- | ---------------- | ------------------------------------------ |\n| Qwen\u002FQwen3.5-35B-A3B           | —       | —          | Text + Vision    | 2,000 RPD total; \u003C=500 RPD\u002Fmodel (dynamic) |\n| Qwen\u002FQwen3.5-27B               | —       | —          | Text             | 2,000 RPD total; \u003C=500 RPD\u002Fmodel (dynamic) |\n| Qwen\u002FQwen-Image                | —       | —          | Image Generation | 2,000 RPD total; model\u002FAIGC-specific caps  |\n| + API-Inference-enabled models | Varies  | Varies     | LLM, MLLM, AIGC  | Dynamic quotas + dynamic concurrency       |\n\n### [NVIDIA NIM](https:\u002F\u002Fbuild.nvidia.com\u002Fexplore\u002Fdiscover) 🇺🇸\n\nFree with NVIDIA Developer Program membership. 100+ models. No daily token cap.\n\nBase URL: `https:\u002F\u002Fintegrate.api.nvidia.com\u002Fv1`\n\n| Model Name                              | Context | Max Output | Modality                               | Rate Limit |\n| --------------------------------------- | ------- | ---------- | -------------------------------------- | ---------- |\n| deepseek-ai\u002Fdeepseek-r1                 | 128K    | ~163K      | Text (reasoning)                       | ~40 RPM    |\n| nvidia\u002Fllama-3.1-nemotron-ultra-253b-v1 | 128K    | 4K         | Text                                   | ~40 RPM    |\n| nvidia\u002Fnemotron-3-super-120b-a12b       | 262K    | 262K       | Text                                   | ~40 RPM    |\n| nvidia\u002Fnemotron-3-nano-30b-a3b          | 128K    | 32K        | Text                                   | ~40 RPM    |\n| meta\u002Fllama-3.1-405b-instruct            | 128K    | 4K         | Text                                   | ~40 RPM    |\n| qwen\u002Fqwen2.5-72b-instruct               | 128K    | 8K         | Text                                   | ~40 RPM    |\n| google\u002Fgemma-4-31b                      | 128K    | 8K         | Text                                   | ~40 RPM    |\n| mistralai\u002Fmistral-large-2-instruct      | 128K    | 4K         | Text                                   | ~40 RPM    |\n| nvidia\u002Fnemotron-nano-2-vl               | 128K    | 8K         | Vision + Text + Video                  | ~40 RPM    |\n| minimax\u002Fminimax-m2.7                    | 128K    | 8K         | Text                                   | ~40 RPM    |\n| + 90 more models                        | Varies  | Varies     | Text, Image, Video, Speech, Embeddings | ~40 RPM    |\n\n### [Ollama Cloud](https:\u002F\u002Follama.com\u002Fsettings\u002Fkeys) 🇺🇸\n\nFree tier with qualitative usage limits. 400+ models from Ollama library. Not OpenAI SDK-compatible; uses [Ollama API](https:\u002F\u002Fdocs.ollama.com\u002Fcloud). [^3]\n\nBase URL: `https:\u002F\u002Fapi.ollama.com`\n\n| Model Name        | Context | Max Output      | Modality         | Rate Limit                          |\n| ----------------- | ------- | --------------- | ---------------- | ----------------------------------- |\n| llama3.1:cloud    | 128K    | Model-dependent | Text             | Session\u002Fweekly limits (unpublished) |\n| deepseek-r1:cloud | 128K    | Model-dependent | Text (reasoning) | Session\u002Fweekly limits (unpublished) |\n| qwen2.5:cloud     | 128K    | Model-dependent | Text             | Session\u002Fweekly limits (unpublished) |\n| gemma2:cloud      | 8K      | Model-dependent | Text             | Session\u002Fweekly limits (unpublished) |\n| mistral:cloud     | 32K     | Model-dependent | Text             | Session\u002Fweekly limits (unpublished) |\n| + 400 more models | Varies  | Varies          | Text             | Session\u002Fweekly limits (unpublished) |\n\n### [OpenRouter](https:\u002F\u002Fopenrouter.ai\u002Fkeys) 🇺🇸\n\n35+ free models (marked with `:free` suffix). OpenAI SDK-compatible. [^4]\n\nBase URL: `https:\u002F\u002Fopenrouter.ai\u002Fapi\u002Fv1`\n\n| Model Name                             | Context | Max Output | Modality         | Rate Limit      |\n| -------------------------------------- | ------- | ---------- | ---------------- | --------------- |\n| deepseek\u002Fdeepseek-r1-0528:free         | 163K    | ~163K      | Text (reasoning) | 20 RPM, 200 RPD |\n| deepseek\u002Fdeepseek-chat-v3-0324:free    | 163K    | 163K       | Text             | 20 RPM, 200 RPD |\n| qwen\u002Fqwen3.6-plus:free                 | 1M      | 65K        | Text             | 20 RPM, 200 RPD |\n| qwen\u002Fqwen3-coder-480b-a35b:free        | 262K    | ~32K       | Text             | 20 RPM, 200 RPD |\n| meta-llama\u002Fllama-4-scout:free          | 10M     | 16K        | Multimodal       | 20 RPM, 200 RPD |\n| meta-llama\u002Fllama-4-maverick:free       | 1M      | 16K        | Multimodal       | 20 RPM, 200 RPD |\n| meta-llama\u002Fllama-3.3-70b-instruct:free | 65K     | ~16K       | Text             | 20 RPM, 200 RPD |\n| google\u002Fgemma-4-31b-it:free             | 256K    | ~8K        | Multimodal       | 20 RPM, 200 RPD |\n| nvidia\u002Fnemotron-3-super-120b-a12b:free | 1M      | ~32K       | Text             | 20 RPM, 200 RPD |\n| openai\u002Fgpt-oss-120b:free               | 131K    | 131K       | Text             | 20 RPM, 200 RPD |\n| minimax\u002Fminimax-m2.5:free              | 196K    | 8K         | Text             | 20 RPM, 200 RPD |\n| mistralai\u002Fdevstral-2512:free           | 256K    | ~32K       | Text             | 20 RPM, 200 RPD |\n| + ~23 more free models                 | Varies  | Varies     | Text \u002F Image     | 20 RPM, 200 RPD |\n\n### [OVHcloud AI Endpoints](https:\u002F\u002Fendpoints.ai.cloud.ovh.net\u002F) 🇫🇷\n\nFree anonymous tier (no API key, no signup): 2 RPM per IP per model. 40+ open-weight models hosted in EU. OpenAI SDK-compatible. [^7]\n\nBase URL: `https:\u002F\u002Foai.endpoints.kepler.ai.cloud.ovh.net\u002Fv1`\n\n| Model Name                    | Context | Max Output | Modality                          | Rate Limit        |\n| ----------------------------- | ------- | ---------- | --------------------------------- | ----------------- |\n| Meta-Llama-3_3-70B-Instruct   | 131K    | ~4K        | Text                              | 2 RPM (anonymous) |\n| Meta-Llama-3_1-8B-Instruct    | 131K    | ~4K        | Text                              | 2 RPM (anonymous) |\n| DeepSeek-R1-Distill-Llama-70B | 131K    | ~32K       | Text (reasoning)                  | 2 RPM (anonymous) |\n| Qwen3-32B                     | 131K    | ~32K       | Text                              | 2 RPM (anonymous) |\n| Qwen3-Coder-30B-A3B-Instruct  | 262K    | ~32K       | Text (code)                       | 2 RPM (anonymous) |\n| Qwen2.5-VL-72B-Instruct       | 128K    | ~8K        | Text + Vision                     | 2 RPM (anonymous) |\n| Mixtral-8x7B-Instruct-v0.1    | 32K     | ~4K        | Text                              | 2 RPM (anonymous) |\n| Mistral-Nemo-Instruct-2407    | 128K    | ~4K        | Text                              | 2 RPM (anonymous) |\n| Qwen3Guard-Gen-8B             | 32K     | ~4K        | Text (safety guard)               | 2 RPM (anonymous) |\n| Qwen3Guard-Gen-0.6B           | 32K     | ~4K        | Text (safety guard)               | 2 RPM (anonymous) |\n| + 30 more models              | Varies  | Varies     | Text, Vision, Code, Image, Speech | 2 RPM (anonymous) |\n\n### [SiliconFlow](https:\u002F\u002Fcloud.siliconflow.cn\u002Faccount\u002Fak) 🇨🇳\n\nFree tier with 14 CNY signup credits. Permanently free models available.\n\nBase URL: `https:\u002F\u002Fapi.siliconflow.cn\u002Fv1`\n\n| Model Name                              | Context | Max Output   | Modality           | Rate Limit         |\n| --------------------------------------- | ------- | ------------ | ------------------ | ------------------ |\n| Qwen\u002FQwen3-8B                           | 131K    | 131K         | Text               | 1,000 RPM, 50K TPM |\n| deepseek-ai\u002FDeepSeek-R1-0528-Qwen3-8B   | ~33K    | 16K          | Text (reasoning)   | 1,000 RPM, 50K TPM |\n| deepseek-ai\u002FDeepSeek-R1-Distill-Qwen-7B | 131K    | Configurable | Text (reasoning)   | 1,000 RPM, 50K TPM |\n| THUDM\u002Fglm-4-9b-chat                     | 32K     | 32K          | Text               | 1,000 RPM, 50K TPM |\n| THUDM\u002FGLM-4.1V-9B-Thinking              | 66K     | 66K          | Vision + Text      | 1,000 RPM, 50K TPM |\n| deepseek-ai\u002FDeepSeek-OCR                | —       | 8K           | Vision (OCR)       | 1,000 RPM, 50K TPM |\n| + embedding\u002Fspeech models               | Varies  | Varies       | Embeddings, Speech | 1,000 RPM, 50K TPM |\n\n## Glossary\n\n| Abbreviation | Meaning             |\n| ------------ | ------------------- |\n| **RPM**      | Requests per minute |\n| **RPD**      | Requests per day    |\n| **TPM**      | Tokens per minute   |\n| **TPD**      | Tokens per day      |\n| **RPS**      | Requests per second |\n\n## Contributing\n\nKnow a free tier that's missing? [Open a PR](contributing.md). Include the provider, endpoint, rate limits (link to their docs), and a few notable models. Trial credits and time-limited promos don't count.\n\n[^1]: Free tier not available in the EU, UK, or Switzerland ([available regions](https:\u002F\u002Fai.google.dev\u002Fgemini-api\u002Fdocs\u002Favailable-regions)).\n[^2]: Groq rate limits vary by model. Llama 4 Maverick is limited to 500 RPD. Most other models get 14,400 RPD ([rate limits](https:\u002F\u002Fconsole.groq.com\u002Fdocs\u002Frate-limits)).\n[^3]: Ollama Cloud measures usage by GPU time, not tokens or requests. Free tier described as \"light usage\" with session limits resetting every 5 hours and weekly limits every 7 days. Pro (50x more) and Max (250x more) plans available. Not OpenAI SDK-compatible; uses the Ollama API.\n[^4]: Free models default to 200 RPD. A one-time purchase of $10+ in credits unlocks 1,000 RPD for free models. OpenRouter also offers a [Free Models Router](https:\u002F\u002Fopenrouter.ai\u002Fdocs\u002Fguides\u002Frouting\u002Frouters\u002Ffree-models-router) (`openrouter\u002Ffree`) and [model fallbacks](https:\u002F\u002Fopenrouter.ai\u002Fdocs\u002Fguides\u002Frouting\u002Fmodel-fallbacks) for chaining models in priority order.\n[^5]: Kilo Code free model list may change over time. nvidia\u002Fnemotron-3-super-120b-a12b:free is for trial use only — prompts are logged by NVIDIA. Auto-router `kilo-auto\u002Ffree` routes to minimax\u002Fminimax-m2.5:free (80%) and stepfun\u002Fstep-3.5-flash:free (20%).\n[^6]: API-Inference is free for registered users. Current published limits are 2,000 requests\u002Fday per user (total across models), with per-model daily quotas dynamically adjusted and capped at 500; concurrency is also dynamically rate-limited. Requires Alibaba Cloud account binding and real-name verification ([limits](https:\u002F\u002Fmodelscope.cn\u002Fdocs\u002Fmodel-service\u002FAPI-Inference\u002Flimits), [intro](https:\u002F\u002Fmodelscope.cn\u002Fdocs\u002Fmodel-service\u002FAPI-Inference\u002Fintro)).\n[^7]: OVHcloud AI Endpoints offers a permanent free anonymous tier (2 requests per minute per IP, per model) with no signup or API key required — click \"Get your free token\" at [endpoints.ai.cloud.ovh.net](https:\u002F\u002Fendpoints.ai.cloud.ovh.net\u002F). Higher rate limits (400 RPM per Public Cloud project per model) require an API key and are billed pay-as-you-go per token; new Public Cloud accounts get up to $200 in free trial credits. Models are hosted in EU data centers.\n","awesome-free-llm-apis 是一个收集了永久免费的大型语言模型（LLM）API密钥的列表。该项目提供了多个知名提供商如Cohere、Google Gemini和Mistral AI等的API信息，每个条目都详细列出了模型名称、上下文长度、最大输出、模态以及速率限制等关键参数，并且大多数API与OpenAI SDK兼容。这使得开发者能够快速找到适合其应用需求的免费LLM资源，适用于初期开发测试或小型项目中对文本生成、图像处理等多种任务的需求探索。",2,"2026-06-11 03:49:56","high_star"]