Awesome Free LLM APIs
Free AI API Providers and Inference Gateways: 2026 Comprehensive Comparison Guide
An authoritative, structured repository of zero-cost AI model endpoints, context limits, modalities, and API rate constraints for developers and researchers.
Direct AI Model Providers
These companies develop proprietary or open-weight models and offer direct API developer keys with free operational tiers.
Cohere 🇨🇦
Base URL: https://api.cohere.com/v2 | Get API Key
Free "Trial" API key, no credit card. 1,000 API calls/month. Non-commercial use only.
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| Command A+ (218B) command-a-plus-05-2026 |
128K | 4K | Text | 20 RPM |
| Command A (111B) command-a-03-2025 |
256K | 4K | Text | 20 RPM |
| Command R+ command-r-plus-08-2024 |
128K | 4K | Text | 20 RPM |
| Command R command-r-08-2024 |
128K | 4K | Text | 20 RPM |
| Command R7B command-r7b-12-2024 |
128K | 4K | Text | 20 RPM |
Google Gemini 🇺🇸
Base URL: https://generativelanguage.googleapis.com/v1beta | Get API Key
Free tier unavailable in EU/UK/Switzerland. Free-tier prompts may be used by Google to improve products. [1]
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| Gemini 3.5 Flash gemini-3.5-flash |
1M | 64K | Text + Image + Audio + Video | 15 RPM, 1,500 RPD |
| Gemini 3.1 Flash-Lite gemini-3.1-flash-lite |
1M | 65K | Text + Image + Audio + Video | 30 RPM, 1,500 RPD |
| Gemini 2.5 Flash gemini-2.5-flash |
1M | 65K | Text + Image + Audio + Video | 15 RPM, 1,500 RPD |
| Gemini 2.5 Pro gemini-2.5-pro |
2M | 65K | Text + Image + Audio + Video | 5 RPM, 50 RPD |
Mistral AI 🇫🇷
Base URL: https://api.mistral.ai/v1 | Get API Key
Free "Experiment" plan, no credit card. ~1B tokens/month. Prompts may be used to improve models.
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| Mistral Medium 3.5 (128B) mistral-medium-2604 |
256K | 256K | Text + Image + Code | ~1 RPS, 500K TPM |
| Mistral Small 4 mistral-small-2603 |
256K | 256K | Text + Image + Code | ~1 RPS, 500K TPM |
| Mistral Large 3 mistral-large-2411 |
256K | 256K | Text | ~1 RPS, 500K TPM |
| Mistral Nemo (12B) open-mistral-nemo |
128K | 128K | Text | ~1 RPS, 500K TPM |
| Codestral codestral-2501 |
256K | 256K | Code | ~1 RPS, 500K TPM |
| Pixtral Large pixtral-large-2411 |
128K | 128K | Text + Image | ~1 RPS, 500K TPM |
Z AI (Zhipu AI) 🇨🇳
Base URL: https://open.bigmodel.cn/api/paas/v4 | Get API Key
Permanent free models, no credit card required.
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| GLM-4.7-Flash glm-4.7-flash |
200K | 128K | Text | 1 concurrent request |
| GLM-4.6V-Flash glm-4.6v-flash |
128K | ~4K | Text + Image | 1 concurrent request |
Multi-Model Inference Gateways
These providers aggregate multiple open architecture models, optimizing processing speed through hardware innovations or token routing networks.
Cerebras 🇺🇸
Base URL: https://api.cerebras.ai/v1 | Explore Gateway
Free tier, no credit card. Ultra-fast inference (~2,600 tok/s). 1M tokens/day cap. 8K context cap on free tier.
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| gpt-oss-120b gpt-oss-120b |
128K (8K on free) | 8K | Text | 30 RPM, 14,400 RPD, 1M TPD |
| zai-glm-4.7 zai-glm-4.7 |
128K (8K on free) | 8K | Text | 10 RPM, 100 RPD, 1M TPD |
Cloudflare Workers AI 🇺🇸
Base URL: https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/run | Explore Gateway
10,000 Neurons/day free. 50+ models available on free tier.
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| @cf/meta/llama-3.3-70b-instruct-fp8-fast | 131K | Shared w/ context | Text | 10K neurons/day (shared) |
| @cf/meta/llama-4-scout-17b-16e-instruct | Up to 10M | Shared w/ context | Multimodal | 10K neurons/day (shared) |
| @cf/openai/gpt-oss-120b | 128K | Shared w/ context | Text | 10K neurons/day (shared) |
| @cf/moonshotai/kimi-k2.7-code | 262K | Shared w/ context | Text (code) | 10K neurons/day (shared) |
| @cf/google/gemma-4-26b-a4b-it | 256K | Shared w/ context | Text | 10K neurons/day (shared) |
| @cf/zhipuai/glm-4.7-flash | 131K | Shared w/ context | Text | 10K neurons/day (shared) |
| @cf/mistralai/mistral-small-3.1-24b-instruct | 128K | Shared w/ context | Text | 10K neurons/day (shared) |
| @cf/deepseek-ai/deepseek-r1-distill-qwen-32b | 32K | Shared w/ context | Text (reasoning) | 10K neurons/day (shared) |
| + 42 more models | Varies | Varies | Text, Image, Audio, Embeddings | 10K neurons/day (shared) |
GitHub Models 🇺🇸
Base URL: https://models.github.ai/inference | Explore Gateway
Free prototyping for all GitHub users. 45+ models. Per-request limits (8K in / 4K out).
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| gpt-5 openai/gpt-5 |
200K | 32K | Text | 10 RPM, 50 RPD |
| gpt-4.1 openai/gpt-4.1 |
1M | 32K | Text | 10 RPM, 50 RPD |
| gpt-4.1-mini openai/gpt-4.1-mini |
1M | 32K | Text | 15 RPM, 150 RPD |
| gpt-4o openai/gpt-4o |
128K | 16K | Text + Vision | 10 RPM, 50 RPD |
| o4-mini openai/o4-mini |
200K | 100K | Text (reasoning) | 10 RPM, 50 RPD |
| Llama-4-Scout-17B-16E meta/Llama-4-Scout-17B-16E |
512K | ~4K | Text + Vision | 15 RPM, 150 RPD |
| Llama-4-Maverick-17B-128E meta/Llama-4-Maverick-17B-128E |
256K | ~4K | Text + Vision | 10 RPM, 50 RPD |
| Meta-Llama-3.3-70B meta/Meta-Llama-3.3-70B |
131K | ~4K | Text | 15 RPM, 150 RPD |
| DeepSeek-R1 deepseek/DeepSeek-R1 |
64K | 8K | Text (reasoning) | 15 RPM, 150 RPD |
| Mistral-Small-3.1 mistralai/Mistral-Small-3.1 |
128K | ~4K | Text + Vision | 15 RPM, 150 RPD |
| + 35 more models | Varies | Varies | Text / Image | Varies by tier |
Groq 🇺🇸
Base URL: https://api.groq.com/openai/v1 | Explore Gateway
Free tier, no credit card. Ultra-fast LPU inference. [2]
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| llama-3.3-70b-versatile | 131K | 32K | Text | 30 RPM, 1,000 RPD |
| llama-3.1-8b-instant | 131K | 131K | Text | 30 RPM, 1,000 RPD |
| llama-4-scout-17b-16e-instruct | 131K | 8K | Text + Vision | 30 RPM, 1,000 RPD |
| qwen3-32b | 131K | 131K | Text | 30 RPM, 1,000 RPD |
| gpt-oss-120b | 131K | 32K | Text | 30 RPM, 1,000 RPD |
Hugging Face 🇺🇸
Base URL: https://router.huggingface.co/v1 | Explore Gateway
100K monthly Inference Provider credits for free users. Routes to Fireworks, Together, Hyperbolic, Nebius, Novita, DeepInfra and others. Thousands of models.
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| Meta-Llama-3.1-8B-Instruct meta-llama/Meta-Llama-3.1-8B-Instruct |
128K | ~4K | Text | Credit-metered |
| Mistral-7B-Instruct-v0.3 mistralai/Mistral-7B-Instruct-v0.3 |
32K | ~4K | Text | Credit-metered |
| Mixtral-8x7B-Instruct-v0.1 mistralai/Mixtral-8x7B-Instruct-v0.1 |
32K | ~4K | Text | Credit-metered |
| Phi-3.5-mini-instruct microsoft/Phi-3.5-mini-instruct |
128K | ~4K | Text | Credit-metered |
| Qwen2.5-7B-Instruct Qwen/Qwen2.5-7B-Instruct |
131K | ~4K | Text | Credit-metered |
| + thousands of community models | Varies | Varies | Text, Image, Audio, Embeddings | 100K credits/month free |
Kilo Code 🇺🇸
Base URL: https://api.kilo.ai/api/gateway | Explore Gateway
Free models with no credit card required. `kilo-auto/free` auto-router routes to minimax/minimax-m2.5:free (80%) and stepfun/step-3.5-flash:free (20%). [5]
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| x-ai/grok-code-fast-1:free | 256K | — | Text (code) | ~200 req/hr |
| minimax/minimax-m2.5:free | 196K | 8K | Text | ~200 req/hr |
| bytedance-seed/dola-seed-2.0-pro:free | — | — | Text | ~200 req/hr |
| nvidia/nemotron-3-super-120b-a12b:free | 262K | 32K | Text | ~200 req/hr |
| arcee-ai/trinity-large-thinking:free | — | — | Text (reasoning) | ~200 req/hr |
| openrouter/free | Varies | Varies | Text | ~200 req/hr |
LLM7.io 🇬🇧
Base URL: https://api.llm7.io/v1 | Explore Gateway
Zero-friction API gateway. No registration needed for basic access. 30+ models. GDPR-compliant.
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| deepseek-r1-0528 | — | — | Text (reasoning) | 30 RPM (120 with token) |
| deepseek-v3-0324 | — | — | Text | 30 RPM (120 with token) |
| gemini-2.5-flash-lite | — | — | Text + Vision | 30 RPM (120 with token) |
| gpt-4o-mini | — | — | Text + Vision | 30 RPM (120 with token) |
| mistral-small-3.1-24b | 32K | — | Text | 30 RPM (120 with token) |
| qwen2.5-coder-32b | — | — | Text (code) | 30 RPM (120 with token) |
| + ~24 more models | Varies | Varies | Text | 30 RPM (120 with token) |
ModelScope 🇨🇳
Base URL: https://api-inference.modelscope.cn/v1 | Explore Gateway
Free API-Inference for registered users. Requires Alibaba Cloud account binding + real-name verification. [6]
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| Qwen/Qwen3.5-35B-A3B | — | — | Text | 2,000 RPD total; <=500 RPD/model |
| Qwen/Qwen3.5-27B | — | — | Text | 2,000 RPD total; <=500 RPD/model |
| + API-Inference-enabled models | Varies | Varies | LLM, MLLM | Dynamic quotas + concurrency |
NVIDIA NIM 🇺🇸
Base URL: https://integrate.api.nvidia.com/v1 | Explore Gateway
Free with NVIDIA Developer Program membership. 100+ models. Rate-limited (no daily token cap).
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| deepseek-ai/deepseek-r1 | 128K | ~163K | Text (reasoning) | ~40 RPM |
| nvidia/nemotron-3-super-120b-a12b | 262K | 262K | Text | ~40 RPM |
| nvidia/nemotron-3-nano-30b-a3b | 128K | 32K | Text | ~40 RPM |
| nvidia/llama-3.1-nemotron-ultra-253b-v1 | 128K | 4K | Text | ~40 RPM |
| meta/llama-3.1-405b-instruct | 128K | 4K | Text | ~40 RPM |
| qwen/qwen2.5-72b-instruct | 128K | 8K | Text | ~40 RPM |
| google/gemma-4-31b | 128K | 8K | Text | ~40 RPM |
| mistralai/mistral-large-2-instruct | 128K | 4K | Text | ~40 RPM |
| minimax/minimax-m2.7 | 128K | 8K | Text | ~40 RPM |
| + 90 more models | Varies | Varies | Text, Image, Video, Speech, Embeddings | ~40 RPM |
Ollama Cloud 🇺🇸
Base URL: https://api.ollama.com | Explore Gateway
Free tier with qualitative usage limits. 400+ models from Ollama library. Not OpenAI SDK-compatible. [3]
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| gpt-oss:120b-cloud | 128K | Model-dependent | Text | Session/weekly limits |
| deepseek-v3.1:671b-cloud | 128K | Model-dependent | Text | Session/weekly limits |
| qwen3-coder:480b-cloud | 128K | Model-dependent | Text (code) | Session/weekly limits |
| kimi-k2:1t-cloud | 262K | Model-dependent | Text | Session/weekly limits |
| glm-4.6:cloud | 128K | Model-dependent | Text | Session/weekly limits |
| deepseek-r1:cloud | 128K | Model-dependent | Text (reasoning) | Session/weekly limits |
| + 30 more cloud models | Varies | Varies | Text | Session/weekly limits |
OpenRouter 🇺🇸
Base URL: https://openrouter.ai/api/v1 | Explore Gateway
~22 free models (marked with `:free` suffix). OpenAI SDK-compatible. [4]
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| qwen/qwen3-coder:free | 1M | 262K | Text (code) | 20 RPM, 200 RPD |
| nvidia/nemotron-3-ultra-550b-a55b:free | 1M | 65K | Text | 20 RPM, 200 RPD |
| nvidia/nemotron-3-super-120b-a12b:free | 1M | 262K | Text | 20 RPM, 200 RPD |
| openai/gpt-oss-120b:free | 131K | 131K | Text | 20 RPM, 200 RPD |
| openai/gpt-oss-20b:free | 131K | 8K | Text | 20 RPM, 200 RPD |
| meta-llama/llama-3.3-70b-instruct:free | 131K | ~16K | Text | 20 RPM, 200 RPD |
| nousresearch/hermes-3-llama-3.1-405b:free | 131K | ~16K | Text | 20 RPM, 200 RPD |
| google/gemma-4-31b-it:free | 262K | 32K | Multimodal | 20 RPM, 200 RPD |
| poolside/laguna-m.1:free | 262K | 32K | Text | 20 RPM, 200 RPD |
| qwen/qwen3-next-80b-a3b-instruct:free | 262K | ~32K | Text | 20 RPM, 200 RPD |
| + ~12 more free models | Varies | Varies | Text / Image | 20 RPM, 200 RPD |
OVHcloud AI Endpoints 🇫🇷
Base URL: https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 | Explore Gateway
Free anonymous tier (no API key, no signup): 2 RPM per IP per model. 20+ open-weight models hosted in EU. OpenAI SDK-compatible. [7]
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| Qwen3.5-397B-A17B | 131K | ~32K | Text | 2 RPM (anonymous) |
| gpt-oss-120b | 128K | ~32K | Text | 2 RPM (anonymous) |
| gpt-oss-20b | 128K | ~8K | Text | 2 RPM (anonymous) |
| Meta-Llama-3_3-70B-Instruct | 131K | ~4K | Text | 2 RPM (anonymous) |
| Llama-3.1-8B-Instruct | 131K | ~4K | Text | 2 RPM (anonymous) |
| Qwen3.6-27B | 131K | ~32K | Text | 2 RPM (anonymous) |
| Qwen3.5-9B | 131K | ~8K | Text | 2 RPM (anonymous) |
| Qwen3-32B | 131K | ~32K | Text | 2 RPM (anonymous) |
| Qwen3-Coder-30B-A3B-Instruct | 262K | ~32K | Text (code) | 2 RPM (anonymous) |
| Qwen2.5-VL-72B-Instruct | 128K | ~8K | Text + Vision | 2 RPM (anonymous) |
| Mistral-Small-3.2-24B-Instruct Mistral-Small-3.2-24B-Instruct-2506 |
128K | ~4K | Text | 2 RPM (anonymous) |
| Mistral-Nemo-Instruct-2407 | 128K | ~4K | Text | 2 RPM (anonymous) |
| Mistral-7B-Instruct-v0.3 | 32K | ~4K | Text | 2 RPM (anonymous) |
SambaNova 🇺🇸
Base URL: https://api.sambanova.ai/v1 | Explore Gateway
Free tier, no credit card. Ultra-fast RDU inference. 20 RPM, 200K tokens/day. [8]
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| DeepSeek-V3.1 | 128K | ~8K | Text | 20 RPM, 20 RPD, 200K TPD |
| DeepSeek-V3.2 (Preview) | 128K | ~8K | Text | 20 RPM, 20 RPD, 200K TPD |
| Meta-Llama-3.3-70B-Instruct | 128K | ~8K | Text | 20 RPM, 20 RPD, 200K TPD |
| gpt-oss-120b | 128K | ~8K | Text | 20 RPM, 20 RPD, 200K TPD |
| MiniMax-M2.7 | 128K | ~8K | Text | 20 RPM, 20 RPD, 200K TPD |
| gemma-4-31B-it (Preview) | 128K | ~8K | Text | 20 RPM, 20 RPD, 200K TPD |
SiliconFlow 🇨🇳
Base URL: https://api.siliconflow.cn/v1 | Explore Gateway
Permanently free models, no credit card required. 200+ paid models also available.
| Model Name & ID | Context | Max Out | Modality | Rate Limits |
|---|---|---|---|---|
| Qwen/Qwen3-8B | 131K | 131K | Text | 30 RPM, 60K TPM |
| deepseek-ai/DeepSeek-R1-Distill-Qwen-7B | 131K | Configurable | Text (reasoning) | 30 RPM, 60K TPM |
Platform Disclaimers & Operational Notes
- Free tier not available in the EU, UK, or Switzerland (available regions).
- Groq rate limits were reduced in 2026. Most models now get 1,000 RPD on the free tier (down from 14,400). Llama 4 Maverick has been deprecated. See rate limits.
- 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.
- Free models default to 200 RPD per model. A one-time purchase of $10+ in credits unlocks 1,000 RPD for free models. OpenRouter also offers a Free Models Router (
openrouter/free) and model fallbacks for chaining models in priority order. Free providers may log prompts for training. - Kilo Code free model list may change over time. nvidia/nemotron-3-super-120b-a12b:free is for trial use only — prompts are logged by NVIDIA. Auto-router
kilo-auto/freeroutes to minimax/minimax-m2.5:free (80%) and stepfun/step-3.5-flash:free (20%). - API-Inference is free for registered users. Current published limits are 2,000 requests/day 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, intro).
- OVHcloud AI Endpoints offers a permanent free anonymous tier (2 requests per minute per IP, per model) with no signup or API key required. 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.
- SambaNova grants $5 in initial credits (valid 30 days) on top of the permanent free tier. The free tier itself persists indefinitely with 20 RPM, 20 RPD, and 200K TPD per model. No credit card required. OpenAI SDK-compatible.
Developer Glossary: API Rate Limit Metrics
- RPM
- Requests per minute
- RPD
- Requests per day
- TPM
- Tokens per minute
- TPD
- Tokens per day
- RPS
- Requests per second
Data compiled for architectural reference. Capabilities, deprecations, and endpoints reflect structural configurations accurate as of mid-2026.
