Free LLM rate-limit calculator
Enter the requests per day and request size you need: see which free LLM tiers cover it alone, which combination of up to three does, and the headroom left.
Published per-minute, per-day and per-month request limits and tokens per minute or per day only, from the same catalog as the free LLM API picker; see how we verify.
Your result
1,000 requests/day at ~800 tokens: 19 free tiers cover it alone
- 1,000 req/day
- ~800 tokens
Covers it alone (19)
Most headroom first. "Cap / day" is what the published limits allow at your request size; the bound that produced it is named.
| Offer | Model | Cap / day | Limited by | Headroom | Spare / day | Card | Status | Notes |
|---|---|---|---|---|---|---|---|---|
 AI21 Labs Free Trial | Jamba LargeOther | 288,000 | requests per minute, sustained | 100% | 287,000 | No card | Verified · Oct 12, 2026 | Foundation model limits: 10 RPS, 200 RPM; same for trial keys |
 Alibaba Cloud Model Studio Free Token Quota | qwen-plusQwen | 864,000 | requests per minute, sustained | 100% | 863,000 | No card | Verified · Oct 12, 2026 | Singapore region account-level limit; other regions differ |
 Cloudflare Workers AI Free Tier | @cf/meta/llama-3.3-70b-instruct-fp8-fastLlama | 432,000 | requests per minute, sustained | 100% | 431,000 | No card | Listed · Oct 21, 2026 | Text Generation: 300 requests per minute; counts against 10,000 free Neurons/day |
 OVHcloud AI Endpoints Free Tier | Qwen3.8-27BQwen | 576,000 | requests per minute, sustained | 100% | 575,000 | Card | Verified · Oct 12, 2026 | 400 req/min per project and per model when authenticated; 2 req/min anonymous |
 Scaleway Generative APIs Free Tier | glm-5.2Other | 360,000 | tokens per minute at your request size | 100% | 359,000 | No card | Verified · Oct 12, 2026 | Supported model; per-model limits in Organization quotas, not on docs |
 LLM7.io Free Tier | DeepSeek-V4-Flash-0731DeepSeek | 86,400 | requests per minute, sustained | 99% | 85,400 | No card | Verified · Oct 12, 2026 | Free token: 100,000 tokens per 24 hours, 60 req/min; limits shared across models |
 Mixedbread Free Starter Credits | Toast 1Other | 144,000 | requests per minute, sustained | 99% | 143,000 | No card | Verified · Oct 12, 2026 | Starter tier: 100 queries/min, 6,000 uploads/min |
 Token Harbor Free Allowance | DeepSeek V4.1 FlashDeepSeek | 86,400 | requests per minute, sustained | 99% | 85,400 | No card | Verified · Oct 12, 2026 | Free plan allowance resets on a rolling 7-day period; no limits published |
| Token Harbor Free Allowance | MiMo V2.6 FlashOther | 86,400 | requests per minute, sustained | 99% | 85,400 | No card | Verified · Oct 12, 2026 | Free allowance resets on a rolling 7-day period; limits unpublished |
 Upstage AI Initiative Free Access | solar-pro4Other | 144,000 | requests per minute, sustained | 99% | 143,000 | No card | Verified · Oct 12, 2026 | Default limits from the chat quickstart table; tier-assigned |
 AMD Radeon Cloud Free APIs | DeepSeek-V4-FlashDeepSeek | 43,200 | requests per minute, sustained | 98% | 42,200 | No card | Verified · Oct 12, 2026 | Typical per-key 30 RPM and 8 concurrent; spend cap per account |
 Sarvam AI Free API Credits | sarvam-105bOther | 57,600 | requests per minute, sustained | 98% | 56,600 | No card | Verified · Oct 12, 2026 | Starter tier chat limit for this model (other chat models 60 req/min) |
 Cohere Free Trial API | command-a-plus-05-2026Other | 28,800 | requests per minute, sustained | 97% | 27,800 | No card | Verified · Oct 12, 2026 | Trial chat limit 20 req/min; trial keys capped at 1,000 API calls a month |
| Agnes 2.5 FlashOther | 21,600 | requests per minute, sustained | 95% | 20,600 | No card | Verified · Oct 12, 2026 | Free plan: 15 req/min, 7M tokens per day shared | |
 Agnes AI Free API | agnes-2.5-flashOther | 14,400 | requests per minute, sustained | 93% | 13,400 | No card | Verified · Oct 12, 2026 | Default keys: text models effective 10 RPM; free keys and Token Plan keys use separate pools |
 NVIDIA NIM Free Inference Endpoints | deepseek-v4.1-flashDeepSeek | 10,000 | requests per day | 90% | 9,000 | No card | Verified · Oct 12, 2026 | Free inference endpoint; limits not published |
 Cerebras Inference Free Trial | gpt-oss-120bGPT | 7,200 | requests per minute, sustained | 86% | 6,200 | Card | Verified · Oct 12, 2026 | Free Trial: 5 RPM, 30K uncached TPM, 1M TPD; 65k context on free vs 131k paid |
 Groq Free Tier | openai/gpt-oss-120bGPT | 1,000 | requests per day | 0% | 0 | No card | Verified · Oct 12, 2026 | Free Plan Limits tab; TPD 200K |
 Infron Free AI Models | deepseek/deepseek-v4.1-flash:freeDeepSeek | 1,000 | requests per day | 0% | 0 | No card | Verified · Oct 12, 2026 | Standard free model; 1,000 requests/day per account shared across free models; requires a $5+ balance |
Best combination falls short
The three largest short tiers from different providers reach 450 requests a day, 45% of what you need.
| Offer | Model | Cap / day | Limited by | Headroom | Spare / day | Card | Status | Notes |
|---|---|---|---|---|---|---|---|---|
 Requesty Free Models | nvidia/nemotron-3-ultra-550b-a55bOther | 200 | requests per day | -400% | -800 (short of 1,000) | No card | Verified · Oct 12, 2026 | Free plan: 200 requests/day across all models |
 Routeway Free Models | DeepSeek V4 Flash (Free)DeepSeek | 200 | requests per day | -400% | -800 (short of 1,000) | No card | Verified · Oct 12, 2026 | Free tier 200 req/day (cannot be increased); 20 RPM per the FAQ |
 OpenRouter Free AI Models | NVIDIA: Nemotron 3 Ultra (free)Other | 50 | requests per day | -1900% | -950 (short of 1,000) | No card | Verified · Oct 12, 2026 | All :free variants: 20 RPM; 50 RPD without credits, 1,000 RPD after 10+ credits |
Not enough alone (4)
Published, but below your need on their own. Largest first.
| Offer | Model | Cap / day | Limited by | Headroom | Spare / day | Card | Status | Notes |
|---|---|---|---|---|---|---|---|---|
| Requesty Free Models | nvidia/nemotron-3-ultra-550b-a55bOther | 200 | requests per day | -400% | -800 (short of 1,000) | No card | Verified · Oct 12, 2026 | Free plan: 200 requests/day across all models |
| Routeway Free Models | DeepSeek V4 Flash (Free)DeepSeek | 200 | requests per day | -400% | -800 (short of 1,000) | No card | Verified · Oct 12, 2026 | Free tier 200 req/day (cannot be increased); 20 RPM per the FAQ |
| OpenRouter Free AI Models | NVIDIA: Nemotron 3 Ultra (free)Other | 50 | requests per day | -1900% | -950 (short of 1,000) | No card | Verified · Oct 12, 2026 | All :free variants: 20 RPM; 50 RPD without credits, 1,000 RPD after 10+ credits |
 OrcaRouter Free AI Models & API Credits | deepseek/deepseek-v4-flash-freeDeepSeek | 50 | requests per day | -1900% | -950 (short of 1,000) | No card | Listed · Sep 10, 2026 | Limits published only at GET /api/free-package/public |
Limits not published (35)
The vendor publishes no requests-per-minute, requests-per-day or tokens-per-minute figure for the free tier, so these are listed, not scored. The offer page links the vendor page.
AkashML $100 Free AI Token CreditsQWEN3.8 27B$100 credits usable on all models; no limits published- Alibaba Cloud Free Trialqwen-plusIndependent free quota per model (typically 1,000,000 tokens), valid 90 days
Alysis Code Free TierDeepSeek V4.1 FlashFree-allowance table; 50 credits every 30 days shared across models
Anthropic AI for Science ProgramClaude Opus 5.5Standard suite; no non-public models; limits follow the org tierAnyAPI Free TierQwen: Qwen3.8 27B (free)Free plan: 100K tokens/day, 'Limited API calls' with no figures
Baseten AI Startup ProgramDeepSeek V4.1 FlashModel API; tool use platform-wide; limits unpublishedBigModel Free GLM Flash ModelsGLM-4.7-FlashFree text model; 128K max output; limits unpublished
BytePlus Free Trialgpt-oss-120bListed under ModelArk free tokens; doc-page limits are not free-tier specific
Claude API New User CreditsClaude Opus 5.5New orgs may start in the Evaluation tier with unpublished limits
DeepInfra DeepStart 50 Billion Free Inference Tokensdeepseek-ai/DeepSeek-V4-Flash-0731DeepStart tokens spend on the whole catalog; featured on the program page; DeepInfra publishes no…
EmberCloud for Startupsglm-5No rate limits published
Experiential Labs Free PlanJevTagged Free on the models page; no rate limits published
fal Startup ProgramNano Banana 2.1Startup-program credits apply to the fal catalog; no per-model limits published; image editing
Fireworks for StartupsGLM-5Build Credits auto-apply to Fireworks usage (offer row); listed in the Fireworks model library;…Free.ai Startup GrantQwen 3.6 35BSelf-hosted open model included in the grant; 100,000 requests per month per key
Google Cloud AI Startup ProgramGemini 3.8 FlashProgram credits cover Gemini on Vertex AI (offer row); listed in the Vertex AI model catalog;…
Google Gemini Developer API Free Tiergemini-3.8-flashFree tier on the pricing page; limits shown only in AI Studio
Helyx AI Free Signup CreditDeepSeek-V4-FlashLive catalog; no per-model daily caps
Hume AI Free TierEVISpeech-to-speech; limits in the API docs
Inception - 100M Free Tokens for Mercury 3Mercury 31M context, up to 131,072 output tokens; 100M free tokens for new accounts; no per-model free…
Mistral AI Free PlanMistral Large 4Free mode limits only in Admin Panel > API > Limits
Nebius Token Factory $1 Trial CreditDeepSeek V4.1 FlashDefaults only shown in the dashboard; the docs example is illustrative
OpenAI for Startupsgpt-6-astraProgram credits are standard API credits; limits follow the org usage tier (program adds upgrades)
OpenCode Zen Free ModelsBig PickleModel id big-pickle; stealth model, data may be used to improve the model
Pollinations Free APIopenai-fastAlias of GPT-5 Nano (openai/gpt-5-nano); priced in Pollen but not paid-only, so Quest Pollen can…
Replicate Try for Freeideogram-ai/ideogram-v3-turboImage generation; free for a limited number of runs
Runware Startup ProgramNano Banana 2.1Image model on the Runware catalog
Sarvam AI Startup ProgramSarvam-105BStartup-program credits cover the Sarvam API stack; limits depend on the account tier; also…
SiliconFlow Free Credits and Free ModelsQwen/Qwen2-7B-InstructOn the free LLM list; limits fixed but unpublished
SmarToken $5 Signup Creditdeepseek-chatCredit consumption varies by route; no rate limits published
Speechmatics $100 Free CreditEnhancedSTT model; Free tier: $100 credit, 2 concurrent real-time sessions
Together AI Startup AcceleratorMiniMaxAI/MiniMax-M3Accelerator credits cover serverless inference; function calling Yes in the serverless table;…
Vercel AI Gateway Free Tierinclusionai/ling-3.1-flashFree-tier-eligible; per-model limits lower than paid and not published as numbers
YepAPI Free $5 CreditNemotron 3 Super$0.00/M input; no hard limits published
Z.AI Free API TierGLM-4.7-FlashPriced Free; concurrency set by the platform per plan
Understand it
A few minutes of play-along, no sign-up.
- Why does the free tier say 429? · Rate limits (RPM, TPM, RPD) · 5 min
- Why does Vietnamese cost more tokens than English? · Tokenizers and languages: why the same sentence costs more · 5 min
Frequently asked questions
- How is "requests per day" computed for a tier?
- From the published limits only: the smallest of requests per day, requests per minute times 1,440 (the per-minute cap sustained all day, a ceiling rather than a promise), tokens per minute times 1,440 divided by your average request size, and, where the vendor publishes them, tokens per day divided by your average request size and requests per month divided by 30. The row says which bound limits you.
- Why is a tier listed under "limits not published"?
- Its vendor publishes no requests-per-minute, requests-per-day or tokens-per-minute figure for the free tier on a page we can cite, so there is nothing to compute from. aifree.dev never guesses a number; the offer page links the vendor page and notes.
- How does the combination work?
- Greedily: the largest tier that falls short goes first, then the next largest from a different provider, up to three providers, until the sum covers your need. Keys from one vendor are never combined because they usually share one quota. It is a spread of load, not a single key.
- What limits are not covered?
- Daily token caps and monthly call caps enter the arithmetic once they are recorded (shown under the model name), but some tiers still carry them only in the notes column, and monthly token caps or credit balances are not modelled. A tier with such a cap in its notes may be overstated; read its notes before relying on it.
- Is the result the same for everyone?
- Yes. The result is a pure function of the catalog and the two numbers in the URL, so a shared link shows the same answer until the next monthly re-check updates a limit.