💡 New here? Every provider below hands you a working key with just an email — no credit card. Pick one and click “Get key →”.
Fastest Path to a Free LLM API Key
Want Gemini?
Google AI Studio — no credit card, instant.
Want fastest inference?
Groq — Llama / Qwen / DeepSeek-R1, no credit card.
Want one key for everything?
OpenRouter — 35+ free models, single endpoint.
Want Claude or GPT-style?
GitHub Models — free for any GitHub user.
Want no rate-limit headaches?
NVIDIA NIM — 40 RPM, no daily token cap.
Free Gemini API Key — 6 models indexed
Google AI Studio gives every developer a free Gemini API key with no credit card. Gemini 2.5 Flash supports 1M context and full multimodal input (text, image, audio, video).
| Provider | Models | Top model (context) | Credit card? | |
|---|
| Google Gemini | 5 | Gemini 2.5 Pro (2.0M) | Not required | Get key → |
| LLM7.io | 1 | gemini-2.5-flash-lite (131K) | Not required | Get key → |
Free OpenAI / GPT API Key — 12 models indexed
OpenAI no longer offers free credits, but GitHub Models (free for any GitHub user) exposes GPT-style endpoints, and OpenRouter, Cerebras, and LLM7.io host gpt-oss open-weight variants you can call with an OpenAI-compatible client.
⚠️ Heads up: There is no official free OpenAI API key in 2026. Anything claiming otherwise is reselling shared keys — use these legitimate free alternatives instead.
| Provider | Models | Top model (context) | Credit card? | |
|---|
| GitHub Models | 4 | gpt-4.1 (1.0M) | Not required | Get key → |
| OpenRouter | 3 | OpenAI: gpt-oss-safeguard-20b (131K) | Not required | Get key → |
| Cerebras | 1 | gpt-oss-120b (128K) | Not required | Get key → |
| Cloudflare Workers AI | 1 | @cf/openai/gpt-oss-120b (128K) | Not required | Get key → |
| LLM7.io | 1 | gpt-4o-mini (131K) | Not required | Get key → |
| Ollama Cloud | 1 | gpt-oss:120b-cloud (128K) | Not required | Get key → |
| OVHcloud AI Endpoints | 1 | gpt-oss-20b (128K) | Not required | Get key → |
Free Claude API Key
Anthropic does not offer a free Claude API tier. The two legitimate paths: (1) use GitHub Models, which proxies a limited Claude allotment to every GitHub user; (2) point Claude Code (cc) at a free backend like Groq or NVIDIA NIM via our config generator.
⚠️ Heads up: If a site offers you "free Claude API keys" with no GitHub or Anthropic login, it is almost always a shared/leaked key that will rate-limit or expire within hours.
No models indexed yet for this brand. Browse all free models →
Free DeepSeek API Key — 18 models indexed
DeepSeek-V3 and DeepSeek-R1 are available across 8+ free providers. Best picks: SiliconFlow (China, lowest latency for CN users), Groq (fastest globally), NVIDIA NIM (no daily token cap), OpenRouter (single key for all variants).
| Provider | Models | Top model (context) | Credit card? | |
|---|
| NVIDIA NIM | 3 | deepseek-ai/deepseek-v4-flash (1.0M) | Not required | Get key → |
| LLM7.io | 2 | deepseek-r1-0528 (131K) | Not required | Get key → |
| Ollama Cloud | 2 | deepseek-v3.1:671b-cloud (128K) | Not required | Get key → |
| SambaNova | 2 | DeepSeek-V3.1 (128K) | Not required | Get key → |
| SiliconFlow | 2 | deepseek-ai/DeepSeek-OCR (131K) | Not required | Get key → |
| DeepSeek | 2 | deepseek-chat (V3.2) (128K) | Not required | Get key → |
| OpenRouter | 1 | DeepSeek: DeepSeek V4 Flash (1.0M) | Not required | Get key → |
| Cloudflare Workers AI | 1 | @cf/deepseek-ai/deepseek-r1-distill-qwen-32b (32K) | Not required | Get key → |
| GitHub Models | 1 | DeepSeek-R1 (64K) | Not required | Get key → |
| Groq | 1 | deepseek-r1-distill-70b (131K) | Not required | Get key → |
| Nscale | 1 | DeepSeek-R1-Distill-Llama-70B (128K) | Not required | Get key → |
Free Llama API Key — 38 models indexed
Meta Llama 3.1 / 3.3 / 4 are the most widely-hosted free open models. Groq has the fastest Llama inference, Cloudflare Workers AI gives 10K req/day free, OpenRouter exposes Llama 3.3 70B with one key.
| Provider | Models | Top model (context) | Credit card? | |
|---|
| NVIDIA NIM | 16 | meta/llama-guard-4-12b (164K) | Not required | Get key → |
| OpenRouter | 5 | Meta: Llama 3.3 70B Instruct (free) (131K) | Not required | Get key → |
| Cloudflare Workers AI | 4 | @cf/meta/llama-4-scout-17b-16e-instruct (10.0M) | Not required | Get key → |
| Groq | 4 | llama-3.3-70b-versatile (131K) | Not required | Get key → |
| GitHub Models | 3 | Llama-4-Scout-17B-16E (512K) | Not required | Get key → |
| OVHcloud AI Endpoints | 2 | Meta-Llama-3_3-70B-Instruct (131K) | Not required | Get key → |
| Nscale | 2 | Llama-3.3-70B-Instruct (128K) | Not required | Get key → |
| Hugging Face | 1 | Meta-Llama-3.1-8B-Instruct (128K) | Not required | Get key → |
| Cerebras | 1 | llama-3.3-70b (128K) | Not required | Get key → |
Free Qwen API Key — 27 models indexed
Alibaba Qwen3 is currently the strongest free coding model. ModelScope and SiliconFlow give Chinese-region users the best latency; OVHcloud and Cerebras serve Europe/US.
| Provider | Models | Top model (context) | Credit card? | |
|---|
| OVHcloud AI Endpoints | 7 | Qwen3-Coder-30B-A3B-Instruct (262K) | Not required | Get key → |
| Alibaba Cloud Model Studio | 4 | Qwen3-Plus (1.0M) | Not required | Get key → |
| ModelScope | 3 | Qwen/Qwen3.5-35B-A3B (131K) | Not required | Get key → |
| OpenRouter | 2 | Qwen: Qwen3 Coder 480B A35B (free) (1.0M) | Not required | Get key → |
| NVIDIA NIM | 2 | qwen/qwen3.5-122b-a10b (262K) | Not required | Get key → |
| Cerebras | 2 | qwen-3-235b-a22b-instruct-2507 (131K) | Not required | Get key → |
| Cloudflare Workers AI | 1 | @cf/deepseek-ai/deepseek-r1-distill-qwen-32b (32K) | Not required | Get key → |
| Groq | 1 | qwen3-32b (131K) | Not required | Get key → |
| Hugging Face | 1 | Qwen2.5-7B-Instruct (131K) | Not required | Get key → |
| LLM7.io | 1 | qwen2.5-coder-32b (131K) | Not required | Get key → |
| Ollama Cloud | 1 | qwen3-coder:480b-cloud (128K) | Not required | Get key → |
| SiliconFlow | 1 | deepseek-ai/DeepSeek-R1-Distill-Qwen-7B (131K) | Not required | Get key → |
| Nebius | 1 | Qwen3-235B-A22B (128K) | Not required | Get key → |
Free Mistral API Key — 19 models indexed
Mistral AI offers a free tier on La Plateforme (~1B tokens/month, no credit card). Mistral models are also mirrored on Cloudflare, NVIDIA NIM, and OVHcloud.
| Provider | Models | Top model (context) | Credit card? | |
|---|
| NVIDIA NIM | 7 | mistralai/codestral-22b-instruct-v0.1 (131K) | Not required | Get key → |
| Mistral AI | 5 | Mistral Medium 3.5 (128B) (256K) | Not required | Get key → |
| OVHcloud AI Endpoints | 2 | Mistral-Small-3.2-24B-Instruct (128K) | Not required | Get key → |
| OpenRouter | 1 | Venice: Uncensored (free) (33K) | Not required | Get key → |
| Cloudflare Workers AI | 1 | @cf/mistralai/mistral-small-3.1-24b-instruct (128K) | Not required | Get key → |
| GitHub Models | 1 | Mistral-Small-3.1 (128K) | Not required | Get key → |
| Hugging Face | 1 | Mistral-7B-Instruct-v0.3 (32K) | Not required | Get key → |
| LLM7.io | 1 | mistral-small-3.1-24b (32K) | Not required | Get key → |
No-Credit-Card Free LLM API Providers
These 24 providers issue an API key with just an email — no card on file, no surprise charges. Together they cover 234+ free models on this site.
OpenRouter
Single API key for 300+ models from every major lab.
Aion Labs
Aion 2.0 — daily token allowance, 131K context, no credit card.
Cohere
Command A, Aya, and embed models — 1,000 API calls/month.
Google Gemini
Google's generous free tier — 1M context, multimodal, 500 RPD.
Mistral AI
Mistral's Experiment plan — ~1B tokens/month, no card needed.
Z AI (Zhipu AI)
Zhipu GLM-4.7-Flash — 200K context, no credit card, 1 concurrent request.
Cerebras
Ultra-fast inference on Cerebras WSE chips — 1M tokens/day.
Cloudflare Workers AI
Edge AI inference — 10,000 neurons/day, 50+ models.
GitHub Models
GPT-4o, o3, Llama 4, DeepSeek-R1 — free for all GitHub users.
Groq
World's fastest LLM inference — ultra-low latency, free tier.
Hugging Face
Hugging Face Inference API — Qwen, Llama, Gemma at ~1,000 RPD.
Kilo Code
Coding-optimized API gateway — routes to the best coding model.
LLM7.io
Free GPT-4o-mini access — rare OpenAI model on a free tier.
ModelScope
Alibaba ModelScope — Qwen3.5, DeepSeek at 2,000 RPD free.
Ollama Cloud
Run Llama, Qwen, Gemma via Ollama API in the cloud.
OVHcloud AI Endpoints
EU-hosted free LLM APIs — GDPR compliant, anonymous tier.
SiliconFlow
China-based free LLM API — DeepSeek, Qwen, BGE at 1,000 RPM.
NVIDIA NIM
100+ open models from NVIDIA — no credit card, 40 RPM.
AI21 Labs
Jamba models with 256K context — 200 RPM, 10 RPS free tier.
Alibaba Cloud Model Studio
Alibaba's official Qwen3 API — 1M context, tiered by region.
DeepSeek
DeepSeek's official API — V3.2 + R1 with 128K context, OpenAI-compatible.
xAI
Grok-4.3, Grok-4.1 Fast, Grok-3 Mini — 1M-2M context, OpenAI-compatible.
Nebius
EU-hosted Llama 3.3 70B + Qwen3 235B — tier-based rate limits.
Nscale
EU-hosted Llama 3.3 70B + Qwen3-Coder + DeepSeek-R1 — fair-use limits.
Every provider on this page exposes an OpenAI-compatible endpoint (or, for Gemini, a wrapper). That means any tool that lets you set a custom baseURL + apiKey works — including Claude Code, Cursor, Codex, Cline, Aider, OpenCode, Continue, and Open WebUI.
Claude Code (cc)
Point Claude Code at Groq, NVIDIA NIM, or OpenRouter — full config snippet.
Cursor
Override Cursor's model with any free OpenAI-compatible endpoint.
Codex CLI
Free backend for OpenAI's Codex CLI tool.
All tools →
Aider, Cline, OpenCode, Continue, Open WebUI, Gemini CLI.
Frequently Asked Questions — Free LLM API Keys
How do I get a free LLM API key in 2026?
The fastest path: sign up for Google AI Studio (Gemini), Groq, or NVIDIA NIM. All three issue an API key instantly with just an email — no credit card. Together they cover 25+ free providers and 234+ models. Pick one from the lists on this page and click "Get key →".
Are free LLM API keys really free, or is there a catch?
Most providers on this page offer a permanently free tier with rate limits (e.g., Groq 14,400 RPD, Gemini 500 RPD, Cerebras 30 RPM). The "catches" to know: (1) some providers train on your free-tier prompts (Gemini does, Groq does not); (2) free tiers often forbid commercial use without an upgrade — Cohere is one example. We label both on each provider page.
Which free LLM API requires no credit card?
234 of our 234 listed models come from providers that issue keys with no credit card on file. The biggest no-card providers are Google AI Studio, Groq, NVIDIA NIM, OpenRouter, Mistral, Cohere, Cerebras, Cloudflare Workers AI, and Z AI.
Is there a free OpenAI API key or free GPT-5 API?
No. OpenAI removed all free credits in 2024 and does not issue free production keys. Any site offering "free OpenAI API keys" is reselling shared/leaked keys that rate-limit within hours. Legitimate free alternatives: GitHub Models (free for every GitHub account), Cerebras (gpt-oss), and OpenRouter (gpt-oss:free).
Is there a free Claude API key?
Anthropic does not offer a free Claude API tier. The only legitimate way to use Claude models for free is GitHub Models (limited monthly Claude budget per GitHub account). If you want the Claude Code CLI experience for free, point it at Groq, NVIDIA NIM, or OpenRouter using our /config/#claude-code generator.
Can I use a free LLM API key for production / commercial use?
Read the provider terms before shipping. Groq, NVIDIA NIM, Cloudflare Workers AI, and Mistral allow commercial use on the free tier with rate limits. Cohere's free trial key is non-commercial only. Google AI Studio free-tier prompts may be used to improve Google products, which is a privacy concern for commercial apps.
Free LLM APIs in 2026: 13 Providers Compared
Last updated: June 2026
If you are building anything with LLMs in 2026, you should not be paying for inference yet. Between Google's Gemini free tier, Groq, Cerebras, OpenRouter, and a dozen smaller providers, you can run real production workloads — chatbots, agents, research pipelines — for $0/month. This is the complete map of every free LLM API still active in June 2026, with rate limits, model access, expiry rules, and which ones do not even ask for a credit card. If you are stacking, this is also the order I would apply in.
Comparison at a glance
| Provider | Models | Free quota | Card | Best for |
|---|
| Google Gemini API | Gemini 2.5 Flash (free tier reduced 2025) | 1,500 req/day Flash, 10 RPM | No | Most accessible free baseline |
| Groq | Llama 3.1, Mixtral, Gemma 2 | 30 RPM, 6K TPM, 1,000 req/day | No | Speed-critical apps (315 TPS) |
| Cerebras | gpt-oss-120b, zai-glm-4.7 | 5 RPM, 30K TPM, 1M tokens/day | No | Very large prompts |
| NVIDIA NIM | Many open + proprietary | Free prototyping tier | Account | Trying new models |
| OpenRouter | Aggregated (50+ models) | Several free-tier models | No | One key, many models |
| OpenAI | GPT-5 family | Data-sharing tokens; ~$5 trial inconsistent | Yes | Eval / one-off tests |
| Anthropic Claude | Opus 4.8, Sonnet 4.6, Haiku 4.5, Fable 5 | ~$5 trial; OSS 6mo Max | Yes | Best-in-class reasoning |
| Mistral La Plateforme | Mistral Small / Large | Free trial credits | Yes | EU compliance |
| Cohere | Command R / R+ | Free trial credits | Yes | RAG-first stacks |
| DeepSeek | DeepSeek V4-Flash / V4-Pro | Cheap pay-as-you-go | Yes | Cheap reasoning |
| xAI Grok | Grok (latest generation) | Limited free credits | Yes | X integration |
| Hugging Face | Open source models | Rate-limited free tier | No | Open weights inference |
| Together AI | 100+ open models | Small free credits | Yes | Open model fine-tuning |
Google Gemini API
- Models: Gemini 2.5 Flash, 2.5 Flash-Lite, and newer 3.x Flash preview models (Gemini 2.0 retired June 2026)
- Free tier: Gemini 2.5 Flash free (Pro and Flash-Lite paid). 1,500 requests/day at 10 RPM. Google reduced free quotas significantly in late 2025 — confirm current limits at ai.google.dev.
- Card required: No. Google account only.
- How to start: ai.google.dev → Get API Key → immediate access
- Best for: Sustained workloads where you want the most "do whatever" capacity without paying
The most accessible free LLM API baseline in 2026 — even after late-2025 quota reductions, Gemini Flash at 1,500 req/day is enough for prototyping. For larger sustained volume, pair Gemini Flash with Cerebras (1M tokens/day) and Groq.
Groq
- Models: Llama 3.1 70B and 8B, Mixtral 8x7B, Gemma 2
- Free tier: 30 requests/minute (Llama 3.1 8B allows 14,400 req/day). Hundreds of tokens/sec on Llama 70B — among the fastest.
- Differentiator: LPUs (Language Processing Units) deliver dramatically faster inference than GPU stacks. Sub-second responses for 70B-class models.
- Card required: No
- How to start: console.groq.com → Sign up → API key
- Best for: Real-time UX (voice, chat with streaming), high throughput batch jobs
If your application is latency-sensitive, Groq's free tier alone is often enough to run production until you hit real scale.
Cerebras
- Models: gpt-oss-120b, zai-glm-4.7 (free-tier roster changes, check the dashboard)
- Free tier: 5 requests per minute, 30K tokens per minute, 1M tokens per day
- Differentiator: Wafer-scale chips designed for inference. Very long context handling and competitive throughput.
- Card required: No
- How to start: cloud.cerebras.ai
- Best for: Long-context tasks (large documents, RAG), batch inference within daily quota
NVIDIA NIM (build.nvidia.com)
- Models: Wide selection — Llama, Mistral, NVIDIA-tuned models, vision models, embeddings
- Free tier: Free during prototyping. Conversion to production usually requires NVIDIA Inception or a paid tier.
- Card required: Account required, no card for free tier
- How to start: build.nvidia.com → sign in → get API key
- Best for: Trying new model architectures before committing to a provider
OpenRouter
- Models: Aggregator routing to 50+ models from every major provider, plus several free-tier models hosted directly
- Free tier: ~28 free models (DeepSeek R1, Llama 3.3 70B, Qwen3 Coder, Gemma 3, more) at 20 RPM and 50 req/day, raised to 1,000/day after a one-time $10 credit purchase
- Card required: No (for free models)
- Pricing: Pay-per-token for paid models, transparent unit economics
- How to start: openrouter.ai
- Best for: Single API key replacing 5+ provider integrations
OpenAI
- Models: GPT-5 family, embeddings
- Free tier: Complimentary tokens if you opt in to share API traffic; the old automatic ~$5 credit is now inconsistent
- Card required: Yes (for any usage beyond initial trial)
- Stacking: OpenAI for Startups program (separate application) gives larger credit grants
- Best for: One-off evaluation. Not for production unless you are a paid user.
Anthropic Claude
- Models: Claude Opus 4.8, Sonnet 4.6, Haiku 4.5, plus the new Fable 5 frontier model
- Free tier: Around $5 in starter credits for new accounts
- Special program: Claude for Open Source (launched February 2026) — qualifying OSS maintainers get six months of Claude Max 20x free, total value $1,200, 10,000 total spots. The largest free Claude grant of 2026.
- Card required: Yes (for API access beyond trial)
- How to start: console.anthropic.com; for OSS apply at the Claude for Open Source program page
- Best for: Highest-quality reasoning when you have OSS standing or don't mind paying
Mistral La Plateforme
- Models: Mistral Small, Mistral Large, Codestral, Embed
- Free tier: Trial credits on signup, modest amount
- Card required: Yes
- Best for: EU-compliant workloads, multilingual generation
Cohere
- Models: Command R, Command R+, Embed, Rerank
- Free tier: Trial credits, generous for evaluation
- Card required: Yes
- Best for: RAG-first applications (Cohere's Rerank is particularly strong)
DeepSeek
- Models: DeepSeek V4-Flash, DeepSeek V4-Pro (thinking + non-thinking, 1M context)
- Free tier: Very cheap pay-as-you-go (V4-Flash from ~$0.14/M input); any free signup grant is small and not guaranteed
- Card required: Yes
- Best for: High-volume reasoning workloads at a fraction of the cost of comparable models
xAI Grok
- Models: Grok (latest generation)
- Free tier: Limited free credits, mostly an evaluation tier
- Card required: Yes
- Best for: Apps integrating with X (Twitter) where Grok's real-time data is the differentiator
Hugging Face Inference API
- Models: Thousands of open source models hosted on the Hub
- Free tier: Rate-limited free access; production use requires Inference Endpoints or PRO subscription
- Card required: No
- Best for: Trying open weights without setting up your own GPU
Together AI
- Models: 100+ open source models (Llama, Mixtral, Qwen, plus fine-tuning support)
- Free tier: Small starter credits
- Card required: Yes
- Best for: Fine-tuning your own model on open weights
How to stack free LLM APIs
If you want to run a real product on $0/month inference, the strategy is:
- Default route through Gemini for general chat and instruction-following work
- Use Groq for real-time voice or interactive flows where latency matters
- Cerebras for long-context RAG or document analysis
- OpenRouter as a safety net — when a primary provider rate-limits you, route the same request through an OpenRouter free model
- Paid burst capacity — keep a small paid balance on OpenAI or Anthropic for queries where you actually need the latest GPT-5 or Claude Opus 4.8 / Fable 5 quality
Most production agents I have built run 90 percent through free tiers and only spend on the top 10 percent of high-stakes calls.
Frequently asked questions
Is there a truly free LLM API I can use without a credit card?
Yes. Google Gemini API, Groq, Cerebras, NVIDIA NIM (build.nvidia.com), and most OpenRouter free-tier models do not require a credit card to start. You sign up with an email or Google account and receive immediate API access with rate limits.
Which free LLM API is the most generous in 2026?
Cerebras leads on raw daily volume — 1M tokens/day with no credit card. Google Gemini API is the most accessible baseline (~1,500 req/day Flash) after the late-2025 reductions. For raw speed, Groq is among the fastest at hundreds of tokens/sec on Llama 70B.
Can I combine multiple free LLM APIs to get more inference?
Yes — and you should. Each provider has independent rate limits, so routing across Gemini + Groq + OpenRouter + Cerebras is a common practice that multiplies your free capacity. Use a router like OpenRouter, LiteLLM, or roll your own with a model registry.
Does OpenAI still give free credits to new accounts?
Not reliably. The old automatic ~$5 new-account credit is now inconsistent; OpenAI's current free route is complimentary tokens for sharing your API traffic. For production, use startup programs (OpenAI for Startups) or stack free tiers like Gemini and Groq.
Is the Claude API free for open source maintainers?
Anthropic's Claude for Open Source program (launched February 2026) gives qualifying OSS maintainers six months of Claude Max 20x free — a $1,200 value with 10,000 total spots. Beyond that, new accounts get a small credit grant (around $5).