Most Models Served Free
Free AI APIs and routers ranked by how many models their free tier or signup credit serves, with the first models named for each provider.
Esta seção está em inglês.
One free key that reaches many models saves you from signing up everywhere: you can try several open and closed models through the same API, compare their answers and switch when one gets rate-limited. Routers and aggregators usually lead this chart; single-lab APIs serve fewer models but often with higher limits. It ranks the verified free tiers and signup credits by how many models they serve free, and names the first few under each value. For model-by-model details such as context size and tool calling, open the model catalog; for the providers with the most generous limits, see highest free rate limits.
Os dados das ofertas são verificados e exibidos em inglês.
| # | Oferta | Models served | Cartão | A pegadinha |
|---|---|---|---|---|
| 1 | OpenRouter Free AI Models | 17 modelsNemotron 3 Ultra, MiniMax M3.1 Flash, Xing4.0 29B A4B +14 more | Não | Free models allow 20 requests/minute and 50/day until you have bought 10 credits (then 1,000/day). A negative balance blocks them, and the lineup rotates (24 variants listed, 17 of them chat models). |
| 2 | OpenCode Zen Free Models | 14 modelsBig Pickle, GLM-5.3-Flash Free, DeepSeek V4.1 Flash Free +11 more | Sim | All free models are limited-time and you must add billing details and credits; Big Pickle, LL Alpha, Exo and MiMo data may be used to improve the models, and NVIDIA trial endpoints log use. |
| 3 | Cloudflare Workers AI Free Tier | 8 models@cf/meta/llama-3.3-70b-instruct-fp8-fast, @cf/openai/gpt-oss-120b, @cf/openai/gpt-oss-20b +5 more | Não | 10,000 Neurons is small (about 49K output tokens a day on Llama 3.3 70B at the published rate); Kimi, GLM-5 and DeepSeek V4 models need Workers Paid or prepaid credits. |
| 4 | AMD Radeon Cloud Free APIs | 7 modelsDeepSeek-V4-Flash, DeepSeek-V4-Flash-Vision-Exp, DeepSeek-V4.1-Flash +4 more | Não | AMD picks which models are served; per-key rate limits, a concurrency allowance, and a spend cap all apply, and only chat-completions and messages endpoints are served. |
| 5 | Google Gemini Developer API Free Tier | 7 modelsgemini-3.8-flash, gemini-3.6-flash, gemini-3.5-flash-lite +4 more | Não | Pro previews, image generation (Nano Banana), Veo video, and Imagen are not available on the free tier; per-model rate limits apply. |
| 6 | LLM7.io Free Tier | 7 modelsDeepSeek-V4-Flash-0731, GLM-5.3-Flash, codestral-latest +4 more | Não | Token usage counts input plus output combined against the daily cap, and per-second/minute/hour request limits apply; a free token is required for the 100,000 tokens/day allowance. |
| 7 | BigModel Free GLM Flash Models | 6 modelsGLM-4.7-Flash, GLM-4.5-Flash, GLM-4-Flash-250414 +3 more | Não | China-facing platform with Chinese-language docs and console; rate limits for the free models are not published on the overview page. |
| 8 | Replicate Try for Free | 6 modelsflux-1.1-pro, flux-dev, flux-kontext-pro +3 more | Não | Free runs are limited (exact cap not published); after hitting it you must add billing and buy credits to continue. |
| 9 | Requesty Free Models | 6 modelsnemotron-3-ultra-550b-a55b, nemotron-3.5-lightning-30b-a3b, gemma-4-31b-it +3 more | Não | Only the 13 designated zero-cost models are free; they are rate-limited per organization and free-model terms can change. |
| 10 | Voyage AI Free Embedding and Reranking Tokens | 6 modelsvoyage-4-large, voyage-4, voyage-4-lite +3 more | Não | Without a payment method you are held to 3 requests and 10K tokens a minute; the published 2,000 RPM limits need a card on file (Usage Tier 1). Older model generations get no free tokens. |
| 11 | OVHcloud AI Endpoints Free Tier | 6 modelsQwen3.8-27B, Qwen3.6-27B, Qwen3.5-397B-A17B +3 more | Sim | Needs a saved payment method (the US site also requires a $0.99 purchase); billing starts automatically once the credit is used up; anonymous API use is capped at 2 requests per minute. |
| 12 | AkashML $100 Free AI Token Credits | 5 modelsQWEN3.8 27B, GLM 5.3, QWEN3.6 35B A3B +2 more | Não | One-time grant for new accounts only, with no renewal; credits are usable only on AkashML's supported open-source models and no expiry is published. |
| 13 | Cohere Free Trial API | 5 modelscommand-a-plus-05-2026, command-a-03-2025, command-a-vision-07-2025 +2 more | Não | Trial API keys are rate limited and not permitted for production or commercial use. |
| 14 | Groq Free Tier | 5 modelsgpt-oss-120b, gpt-oss-20b, qwen3.8-27b +2 more | Não | Rate limits apply at the organization level and are tight (e.g. 30 RPM and 1K RPD on many models); the docs caution limits may change and exceptions exist. |
| 15 | Helyx AI Free Signup Credit | 5 modelsDeepSeek-V4-Flash, DeepSeek-V4-Pro, gemini-3.1-flash-lite +2 more | Não | The $8 portion is held until you contact support (an anti-abuse measure); top-ups are via crypto. |
| 16 | Scaleway Generative APIs Free Tier | 5 modelsglm-5.2, deepseek-v4-flash-0731, qwen3.8-27b +2 more | Não | Free Tier applies to your most expensive tokens first; once exceeded, per-token billing starts with a 1,000-token minimum billing unit. |
| 17 | SmarToken $5 Signup Credit | 5 modelsdeepseek-chat, kimi-k2, qwen-plus +2 more | Não | The $5 is promotional balance that cannot be withdrawn as cash, granted once per eligible account, and new Google signups must pass risk checks before 31 Dec 2026 (UTC+8). |
| 18 | AgentRouter Free Credits | 4 modelsgpt-6-astra, gpt-6.1-sol, claude-opus-5-5 +1 more | Não | Claude and GPT models come in limited daily batches that run out; only supported coding clients are allowed, and misuse moves accounts to a probation group with doubled rates. |
| 19 | Agnes AI Free API | 4 modelsagnes-2.5-flash, agnes-3.0-flash, agnes-image-2.1-flash +1 more | Não | Free video pricing is a limited-time promotion; free access is rate-limited (about 20 effective RPM text, 1 RPM video) and the docs say all limits may be adjusted in the future. |
| 20 | Alysis Code Free Tier | 4 modelsDeepSeek V4.1 Flash, GLM 5.3 Flash, GPT-6 Luna +1 more | Não | Unused credits do not carry over between 30-day periods; the free tier now spans multiple providers (DeepSeek, GLM, GPT-6 Luna, Claude Sonnet) and is subject to fair-use limits. |
Como este ranking é ordenado
Verified offers of type free tier, fully free, free credits or free trial (startup and student programs are left out because they need an application) whose model list has at least two models. Sorted by the number of distinct models on that list, most first; the same model under a router tag such as ":free" counts once. Ties go to offers that need no credit card, then by name.
As linhas vêm do mesmo catálogo das páginas de ofertas, verificado de novo todo mês. Os valores são os que cada provedor informa; o aifree.dev nunca converte créditos ou tokens em dólares. Como verificamos (em inglês)
Perguntas frequentes
Esta seção está em inglês.
- Which free AI API gives access to the most models?
- The first row of the chart, right now. Model routers and aggregators usually serve the most, because they pass requests on to many labs; their free models can change often, so check the offer page before you depend on one.
- Are all of these models free without limits?
- No. The models are free within the offer rules, usually a shared daily request cap or a credit pool. More models does not mean more free use; the limits are on each offer page.
- Why is a big provider missing or ranked low?
- Only models that the offer records as served free are counted, and offers that need an application, such as startup programs, are left out. A provider that serves many models only on paid plans ranks by its free list alone.