Free LLM All free tiers / Requesty
Requesty free tier
Free plan: the free models can be used at no charge without a payment method, up to 200 requests per day across all of them. OpenAI-compatible API at https://router.requesty.ai/v1 (EU endpoint: router.eu.requesty.ai).
Kind: Free models · Provider type: Router
Card needed: no (stated by the provider)
Limits: New organizations: 200 requests/day and 20 requests/min, shared across all free models combined; paying organizations: 1,000/day and 60/min. Requesty may change the limit or the set of free models.
Your data: Split per model: most free models are labelled "Training Permitted": their provider keeps and trains on your prompts, and on the free plan Requesty may also use them to improve its service (Terms 6.2(a)). The others are not used for training. Prompt logging is on by default (kept up to 30 days) and can be switched off per API key.
Free models and offers (13)
| Model | Requests | Tokens | Other limits / notes |
|---|---|---|---|
| google/gemma-4-31b-it | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Not used for training; provider keeps data 30 days. 262,144 token context. |
| mistral/leanstral-1-5 | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Not used for training; provider keeps data 0 days (zero retention). 262,144 token context. Hosted in the EU. |
| novita/inclusionai/ling-3.0-tiny | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Not used for training; provider keeps data 0 days (zero retention). 262,144 token context. |
| novita/ling-3.1-flash | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Not used for training; provider keeps data 0 days (zero retention). 262,144 token context. |
| nvidia/muse-glimmer-30b | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Training Permitted: the provider keeps prompts and trains on them; on the free plan Requesty may also use them (Terms 6.2(a)). 131,072 token context. |
| nvidia/nemotron-3-nano-30b-a3b | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Training Permitted: the provider keeps prompts and trains on them; on the free plan Requesty may also use them (Terms 6.2(a)). 262,144 token context. |
| nvidia/nemotron-3-nano-omni-30b-a3b-reasoning | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Training Permitted: the provider keeps prompts and trains on them; on the free plan Requesty may also use them (Terms 6.2(a)). 131,072 token context. |
| nvidia/nemotron-3-super-120b-a12b | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Training Permitted: the provider keeps prompts and trains on them; on the free plan Requesty may also use them (Terms 6.2(a)). 1,048,576 token context. |
| nvidia/nemotron-3-ultra-550b-a55b | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Training Permitted: the provider keeps prompts and trains on them; on the free plan Requesty may also use them (Terms 6.2(a)). 1,048,576 token context. |
| nvidia/nemotron-3.5-content-safety | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Training Permitted: the provider keeps prompts and trains on them; on the free plan Requesty may also use them (Terms 6.2(a)). 131,072 token context. |
| nvidia/nemotron-3.5-lightning-30b-a3b | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Training Permitted: the provider keeps prompts and trains on them; on the free plan Requesty may also use them (Terms 6.2(a)). 1,048,576 token context. |
| poolside/laguna-m.1 | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Training Permitted: the provider keeps prompts and trains on them; on the free plan Requesty may also use them (Terms 6.2(a)). 32,768 token context. |
| poolside/laguna-xs.2 | 20/min · 200/day | n/p | Shared across all free models; 1,000/day and 60/min for paying orgs · Training Permitted: the provider keeps prompts and trains on them; on the free plan Requesty may also use them (Terms 6.2(a)). 32,768 token context. |
n/p = not published on the provider's page. Used for training: yes.
Notes
- Only one account per person or legal entity is allowed; extra accounts to get around limits are banned (Terms 3.1).
- Requesty says these models are free for now and pricing may change, announced in its changelog ahead of time.
- The list of free models and their training flags comes from Requesty's public models API.
Sources
- Requesty free models checked 2026-11-04
- Requesty Terms of Service (3.1, 4.6, 5.9, 6.2(a), 6.3) checked 2026-11-04
- Requesty data privacy checked 2026-11-04
- Requesty models API checked 2026-11-04
Need more than the free tier? Compare paid per-token prices on prices.branolabs.com.