Directory of Free LLM APIs: Compare 160+ Models
Showing 160 of 160 free LLM models
Discover and filter 160+ free LLM models across 29 providers. Find APIs by capability (vision, reasoning), rate limits, or no-credit-card requirements, and get the perfect free AI model for your project.
51 models verified via live API · refreshed Jun 28, 2026 — how we verify
| Provider | Model | Score | Context | Modality | Rate Limit | Status |
|---|---|---|---|---|---|---|
| MiniMax: MiniMax M3 Paid MiniMax: MiniMax M3 | 94 | 1.0M | 200 req/day (free tier) | Online | ||
| Nex AGI: Nex-N2-Pro Paid Nex AGI: Nex-N2-Pro | 91 | 262K | 200 req/day (free tier) | Online | ||
| Gemini 3.5 Flash | 90 | 1.0M | 15 RPM, 1,500 RPD | Online | ||
| minimaxai/minimax-m3 Verified MiniMax: MiniMax M3 | 88 | 1.0M | Up to 40 RPM | Online | ||
| DeepSeek: DeepSeek V4 Flash | 88 | 1.0M | 200 req/day (free tier) | Online | ||
| NVIDIA: Nemotron 3 Ultra (free) Verified NVIDIA: Nemotron 3 Ultra (free) | 88 | 1.0M | 200 req/day (free tier) | Online | ||
| MoonshotAI: Kimi K2.6 | 87 | 262K | 200 req/day (free tier) | Online | ||
| deepseek-ai/deepseek-v4-pro Verified deepseek-ai/deepseek-v4-pro | 85 | 1.0M | Up to 40 RPM | Online | ||
| Z.ai: GLM 5.1 Paid | 85 | 203K | 200 req/day (free tier) | Online | ||
| deepseek-ai/deepseek-v4-flash Verified DeepSeek: DeepSeek V4 Flash | 82 | 1.0M | Up to 40 RPM | Online | ||
| moonshotai/kimi-k2.6 Verified MoonshotAI: Kimi K2.6 | 82 | 262K | Up to 40 RPM | Online | ||
| Qwen3.6-27B | 82 | 131K | 2 RPM (anonymous) | Online | ||
| 80 | 262K | 200 req/day (free tier) | Online | |||
| z-ai/glm-5.1 Verified | 80 | 203K | Up to 40 RPM | Online | ||
| Cohere: North Mini Code (free) Verified Cohere: North Mini Code (free) | 80 | 256K | 200 req/day (free tier) | Online | ||
| minimaxai/minimax-m2.7 Verified MiniMax: MiniMax M3 | 76 | 205K | Up to 40 RPM | Online | ||
| Google: Gemma 4 31B (free) Verified Google: Gemma 4 31B (free) | 76 | 262K | 200 req/day (free tier) | Online | ||
| MiniMax: MiniMax M3 | 75 | 205K | 200 req/day (free tier) | Online | ||
| stepfun-ai/step-3.7-flash Verified | 75 | 256K | Up to 40 RPM | Online | ||
| Qwen/Qwen3.5-27B | 74 | 131K | 2,000 RPD total; <=500 RPD/model (dynamic) | Online | ||
| 73 | 262K | 200 req/day (free tier) | Online | |||
| Google: Gemma 4 26B A4B (free) Verified Google: Gemma 4 26B A4B (free) | 73 | 262K | 200 req/day (free tier) | Online | ||
| Qwen3.5-9B | 73 | 131K | 2 RPM (anonymous) | Online | ||
| MiniMax-M2.7 | 73 | 128K | 20 RPM, 20 RPD, 200K TPD | Online | ||
| Qwen3.5-397B-A17B | 73 | 131K | 2 RPM (anonymous) | Online | ||
| MiniMax: MiniMax M3 | 72 | 196K | ~200 req/hr | Online | ||
| NVIDIA: Nemotron 3 Nano Omni (free) Verified NVIDIA: Nemotron 3 Nano Omni (free) | 72 | 256K | 200 req/day (free tier) | Online | ||
| NVIDIA: Nemotron 3 Super (free) Verified NVIDIA: Nemotron 3 Super (free) | 72 | 1.0M | 200 req/day (free tier) | Online | ||
| qwen/qwen3.5-397b-a17b Verified Qwen3.5-397B-A17B | 70 | 256K | Up to 40 RPM | Online | ||
| qwen/qwen3.5-122b-a10b Verified qwen/qwen3.5-122b-a10b | 70 | 262K | Up to 40 RPM | Online | ||
| Qwen/Qwen3.5-35B-A3B | 70 | 131K | 2,000 RPD total; <=500 RPD/model (dynamic) | Online | ||
| Owl Alpha Verified | 69 | 1.0M | 200 req/day (free tier) | Online | ||
| 69 | 131K | ~200 req/hr | Online | |||
| Poolside: Laguna XS.2 (free) Verified | 68 | 262K | 200 req/day (free tier) | Online | ||
| Poolside: Laguna M.1 (free) Verified | 68 | 262K | 200 req/day (free tier) | Online | ||
| NVIDIA: Nemotron 3 Super (free) | 68 | 262K | ~200 req/hr | Online | ||
| Google: Gemma 4 26B A4B (free) | 66 | 256K | 10K neurons/day (shared) | Online | ||
| 65 | 128K | Session/weekly limits (unpublished) | Online | |||
| MoonshotAI: Kimi K2.6 | 64 | 262K | 10K neurons/day (shared) | Online | ||
| GLM-4.7-Flash | 64 | 200K | 1 concurrent request | Online | ||
| stepfun-ai/step-3.5-flash Verified | 63 | 262K | Up to 40 RPM | Online | ||
| GLM-4.6V-Flash | 63 | 128K | 1 concurrent request | Online | ||
| Gemini 3.1 Flash-Lite | 62 | 1.0M | 30 RPM, 1,500 RPD | Online | ||
| o4-mini | 62 | 200K | 10 RPM, 50 RPD | Online | ||
| 61 | 256K | ~1 RPS, 500K TPM | Online | |||
| 61 | 131K | ~200 req/hr | Online | |||
| zai-glm-4.7 | 61 | 128K | 10 RPM, 100 RPD, 1M TPD | Online | ||
| deepseek-r1:cloud | 61 | 128K | Session/weekly limits (unpublished) | Online | ||
| 60 | 128K | Session/weekly limits (unpublished) | Online | |||
| Qwen: Qwen3 Coder 480B A35B (free) Verified Qwen: Qwen3 Coder 480B A35B (free) | 60 | 1.0M | 200 req/day (free tier) | Online | ||
| Nous: Hermes 3 405B Instruct (free) Verified | 59 | 131K | 200 req/day (free tier) | Online | ||
| 59 | 128K | 20 RPM, 20 RPD, 200K TPD | Online | |||
| 59 | 256K | See provider page | Online | |||
| OpenAI: gpt-oss-120b (free) Verified OpenAI: gpt-oss-120b (free) | 59 | 131K | 200 req/day (free tier) | Online | ||
| 58 | 131K | See provider page | Online | |||
| 58 | 131K | See provider page | Online | |||
| gpt-4.1 | 58 | 1.0M | 10 RPM, 50 RPD | Online | ||
| NVIDIA: Nemotron 3.5 Content Safety (free) | 58 | 128K | 200 req/day (free tier) | Online | ||
| gemma-4-31B-it (Preview) | 57 | 128K | 20 RPM, 20 RPD, 200K TPD | Online | ||
| 56 | 128K | 10K neurons/day (shared) | Online | |||
| 56 | 256K | ~200 req/hr | Online | |||
| gpt-4.1-mini | 56 | 1.0M | 15 RPM, 150 RPD | Online | ||
| Qwen: Qwen3 Next 80B A3B Instruct (free) | 56 | 262K | 200 req/day (free tier) | Online | ||
| Google: Lyria 3 Pro Preview Verified Google: Lyria 3 Pro Preview | 56 | 1.0M | 200 req/day (free tier) | Online | ||
| Google: Lyria 3 Clip Preview Verified Google: Lyria 3 Clip Preview | 56 | 1.0M | 200 req/day (free tier) | Online | ||
| @cf/zhipuai/glm-4.7-flash | 56 | 131K | 10K neurons/day (shared) | Online | ||
| gpt-5 | 56 | 200K | 10 RPM, 50 RPD | Online | ||
| deepseek-r1-0528 | 56 | 131K | 30 RPM (120 with token) | Online | ||
| gpt-oss:120b-cloud | 56 | 128K | Session/weekly limits (unpublished) | Online | ||
| NVIDIA: Nemotron 3 Nano 30B A3B (free) | 55 | 256K | 200 req/day (free tier) | Online | ||
| Qwen3-Coder-30B-A3B-Instruct | 55 | 262K | 2 RPM (anonymous) | Online | ||
| 54 | 262K | Session/weekly limits (unpublished) | Online | |||
| gpt-oss:120b-cloud | 54 | 128K | 30 RPM, 14,400 RPD, 1M TPD | Online | ||
| Z.ai: GLM 4.5 Air Paid Z.ai: GLM 4.5 Air | 54 | 131K | 200 req/day (free tier) | Online | ||
| GPT-OSS 120B Verified | 53 | 131K | See provider page | Online | ||
| NVIDIA: Nemotron Nano 12B 2 VL (free) | 53 | 128K | 200 req/day (free tier) | Online | ||
| deepseek-v3-0324 | 53 | 131K | 30 RPM (120 with token) | Online | ||
| DeepSeek-V3.1 | 53 | 128K | 20 RPM, 20 RPD, 200K TPD | Online | ||
| 52 | 256K | ~1 RPS, 500K TPM | Online | |||
| OpenAI: gpt-oss-20b (free) Verified | 52 | 131K | 200 req/day (free tier) | Online | ||
| Venice: Uncensored (free) Verified | 52 | 33K | 200 req/day (free tier) | Online | ||
| zai-glm-4.7 | 52 | 128K | Session/weekly limits (unpublished) | Online | ||
| nvidia/nemotron-3.5-content-safety Verified NVIDIA: Nemotron 3.5 Content Safety (free) | 52 | 128K | Up to 40 RPM | Online | ||
| Gemini 2.5 Flash | 52 | 1.0M | 15 RPM, 1,500 RPD | Online | ||
| 51 | 10.0M | 10K neurons/day (shared) | Online | |||
| NVIDIA: Nemotron Nano 9B V2 (free) Verified NVIDIA: Nemotron Nano 9B V2 (free) | 51 | 128K | 200 req/day (free tier) | Online | ||
| Gemini 2.5 Pro | 51 | 2.0M | 5 RPM, 50 RPD | Online | ||
| Llama-4-Scout-17B-16E | 51 | 512K | 15 RPM, 150 RPD | Online | ||
| Llama-4-Scout-17B-16E | 51 | 256K | 10 RPM, 50 RPD | Online | ||
| gpt-4o | 51 | 128K | 10 RPM, 50 RPD | Online | ||
| 50 | 33K | 200 req/day (free tier) | Online | |||
| Llama-4-Scout-17B-16E | 50 | 131K | 30 RPM, 1,000 RPD | Online | ||
| Aion 2.0 | 50 | 128K | 15 RPM, 20K TPD | Online | ||
| gemini-2.5-flash-lite | 50 | 131K | 30 RPM (120 with token) | Online | ||
| Qwen2.5-VL-72B-Instruct | 50 | 128K | 2 RPM (anonymous) | Online | ||
| Mistral-Small-3.2-24B-Instruct | 50 | 128K | 2 RPM (anonymous) | Online | ||
| Mistral-Nemo-Instruct-2407 | 50 | 128K | 2 RPM (anonymous) | Online | ||
| Meta: Llama 3.3 70B Instruct (free) Verified Meta: Llama 3.3 70B Instruct (free) | 50 | 131K | 200 req/day (free tier) | Online | ||
| 49 | 256K | ~1 RPS, 500K TPM | Online | |||
| 49 | 256K | ~1 RPS, 500K TPM | Online | |||
| 49 | 33K | 200 req/day (free tier) | Online | |||
| Free Models Router Verified | 49 | 200K | 200 req/day (free tier) | Online | ||
| gpt-oss-20b | 49 | 128K | 2 RPM (anonymous) | Online | ||
| qwen3-32b | 49 | 131K | 30 RPM, 1,000 RPD | Online | ||
| mistral-small-3.1-24b | 49 | 32K | 30 RPM (120 with token) | Online | ||
| 48 | 33K | See provider page | Online | |||
| DeepSeek-R1 | 48 | 64K | 15 RPM, 150 RPD | Online | ||
| 47 | 33K | Unlimited for free models | Online | |||
| 47 | 33K | See provider page | Online | |||
| Meta: Llama 3.3 70B Instruct (free) | 47 | 131K | 15 RPM, 150 RPD | Online | ||
| Meta: Llama 3.3 70B Instruct (free) | 47 | 131K | 2 RPM (anonymous) | Online | ||
| Meta: Llama 3.3 70B Instruct (free) | 47 | 131K | 30 RPM, 1,000 RPD | Online | ||
| nvidia/llama-3.3-nemotron-super-49b-v1.5 | 47 | 131K | Up to 40 RPM | Online | ||
| GPT-OSS 20B Verified | 46 | 131K | See provider page | Online | ||
| 46 | 131K | See provider page | Online | |||
| 45 | 128K | 15 RPM, 150 RPD | Online | |||
| 45 | 32K | Credit-metered | Online | |||
| mistral-small-3.1-24b | 45 | 128K | 10K neurons/day (shared) | Online | ||
| Meta: Llama 3.2 3B Instruct (free) Verified Meta: Llama 3.2 3B Instruct (free) | 45 | 131K | 200 req/day (free tier) | Online | ||
| 44 | 256K | 20 RPM | Online | |||
| 44 | 131K | 30 RPM, 60K TPM | Online | |||
| 44 | 128K | Credit-metered | Online | |||
| 44 | 33K | See provider page | Online | |||
| 44 | 33K | See provider page | Online | |||
| 44 | 128K | 20 RPM | Online | |||
| 44 | 128K | 20 RPM | Online | |||
| 44 | 128K | 20 RPM | Online | |||
| 44 | 131K | 200 req/day (free tier) | Online | |||
| Nemotron 3 Ultra 550B A55B Verified | 44 | 1.0M | Online | |||
| 44 | 128K | 15 RPM, 20K TPD | Online | |||
| Mistral-Nemo-Instruct-2407 | 44 | 128K | ~1 RPS, 500K TPM | Online | ||
| Llama-3.1-8B-Instruct | 44 | 131K | 2 RPM (anonymous) | Online | ||
| Llama-3.1-8B-Instruct | 44 | 131K | 30 RPM, 1,000 RPD | Online | ||
| Phi-4 | 44 | 131K | See provider page | Online | ||
| qwen2.5-coder-32b | 44 | 131K | 30 RPM (120 with token) | Online | ||
| gpt-4o-mini | 44 | 131K | 30 RPM (120 with token) | Online | ||
| 43 | 32K | Credit-metered | Online | |||
| Meta: Llama 3.3 70B Instruct (free) | 43 | 131K | 10K neurons/day (shared) | Online | ||
| DeepSeek-R1 | 43 | 131K | Community-powered, no hard cap | Online | ||
| meta/llama-3.1-70b-instruct Verified meta/llama-3.1-70b-instruct | 43 | 131K | Up to 40 RPM | Online | ||
| 41 | 32K | 15 RPM, 20K TPD | Online | |||
| meta/llama-3.1-70b-instruct | 41 | 131K | See provider page | Online | ||
| @cf/deepseek-ai/deepseek-r1-distill-qwen-32b | 41 | 32K | 10K neurons/day (shared) | Online | ||
| 40 | 131K | $25/month free credits, resets monthly | Online | |||
| DeepSeek V4 Flash Verified | 40 | 1.0M | Online | |||
| meta/llama-3.1-70b-instruct | 40 | 131K | Community-powered, no hard cap | Online | ||
| Qwen2.5-7B-Instruct | 40 | 131K | Credit-metered | Online | ||
| meta/llama-3.2-11b-vision-instruct Verified meta/llama-3.2-11b-vision-instruct | 40 | 131K | Up to 40 RPM | Online | ||
| Pixtral Large | 40 | 128K | ~1 RPS, 500K TPM | Online | ||
| meta/llama-3.2-3b-instruct Verified Meta: Llama 3.2 3B Instruct (free) | 39 | 131K | Up to 40 RPM | Online | ||
| meta/llama-3.1-70b-instruct | 39 | 131K | Unlimited for free models | Online | ||
| Mistral Large (24.11) | 39 | 131K | See provider page | Online | ||
| GPT-OSS Safeguard 20B Verified | 38 | 131K | See provider page | Online | ||
| meta/llama-3.2-1b-instruct Verified meta/llama-3.2-1b-instruct | 38 | 131K | Up to 40 RPM | Online | ||
| MiMo-V2.5 Verified | 37 | 1.0M | Online | |||
| Llama-3.1-8B-Instruct | 36 | 128K | Credit-metered | Online | ||
| North Mini Code Verified | 34 | 256K | Online | |||
| meta/llama-guard-4-12b Verified meta/llama-guard-4-12b | 34 | 164K | Up to 40 RPM | Online | ||
| big-pickle Verified | 33 | N/A | Online | |||
| 32 | 131K | $25/month free credits, resets monthly | Online |
How to Get Started with Free LLM APIs
- Pick a free LLM model — Click any model name to see details, rate limits, and API key signup link.
- Get your API key — Sign up on the provider's website (most require no credit card).
- Copy the config — Go to the Config Generator, pick your tool and backend, copy the ready-to-use snippet.
- Test it — Use the Playground to test your API key before integrating.
New to LLM terminology? Check the 📖 Glossary — 22 terms explained in plain English →
FAQ: Common questions about free LLM APIs →About This Free LLM API Directory
Finding reliable free LLM API resources online can be frustrating. Many developers traditionally rely on static GitHub repositories to find endpoints. While those lists are a good starting point, they often become outdated quickly, leaving you with dead links, expired API keys, and unverified rate limits.
That's why we built this dynamic, auto-updating directory. If you are looking for a reliable alternative to GitHub free LLM API lists, this page tracks over 160 free LLM models online in real-time. Whether you need a free API key for text generation, vision, or coding tasks, you can compare context windows, capabilities, and strict rate limit data side-by-side.
Our goal is to be the most accurate and comprehensive list of free AI APIs for developers. Use the filters above to find providers that don't require credit cards or phone verification, and grab your free API keys to start building immediately.