Home / Inference
Requesty
docs.requesty.ai/features/free-modelsInference gateway, free modelsinference/requestyJSONcomparereport an errorprovider JSON
- Free allowance
- Free plan: access to all free models at no token charge, 200 requests per day and 20 requests per minute for a new organization, counted across all free models combined; free models include nvidia/nemotron-3-ultra-550b-a55b, nvidia/nemotron-3-super-120b-a12b, google/gemma-4-31b-it and...
- Verdict
- caveatusable, with a named limit or an unprobed read path
- Cost
- free tier
- Identity
- emailno card, no phone
- Binding limit
- request-ratenew organizations: 200 requests per day and 20 requests per minute across all free models combined; paying organizations: 1,000 per day and 60 per minute; daily limit resets daily
- Evidence
- documented6 sources; a first-party page was read and the sentence quoted
- Freshness
- stale on 2026-11-29checked 2026-10-29 (live); stale on 2026-11-29
| Free tier | Free plan: access to all free models at no token charge, 200 requests per day and 20 requests per minute for a new organization, counted across all free models combined; free models include nvidia/nemotron-3-ultra-550b-a55b, nvidia/nemotron-3-super-120b-a12b, google/gemma-4-31b-it and poolside/laguna-m.1 unit requests; metered on request-count; the docs mark the models 'free for now' and announce any price change in the changelog ahead of time; paying organizations get 1,000 requests per day and 60 per minute |
|---|
| Price | none stated |
|---|
| Limits | - request-rate (binding) new organizations: 200 requests per day and 20 requests per minute across all free models combined; paying organizations: 1,000 per day and 60 per minute; daily limit resets daily source
|
|---|
| Binding | request-rate the 200 requests per day budget is shared across all free models and binds long before any per-minute figure |
|---|
| Identity | email no credit card required for the free models, quoted from the pricing and FAQ pages; the sign-up form at app.requesty.ai was not captured, so any further check on it is not established |
|---|
| Egress | not published |
|---|
| Interface | rest |
|---|
| Tooling | any OpenAI-compatible client at https://router.requesty.ai/v1 with the model id written provider/model; an Anthropic-style /anthropic/v1/messages path also exists |
|---|
| Composes with | the free models route through the same gateway as the 600+ paid ones, so a fallback policy can move to a paid model when the daily limit is hit |
|---|
| Reliability | free models are labelled free for now and the limit is shared across all of them, so one busy model exhausts the budget for the others |
|---|
| Notes | per-model limits do not exist: the figures are per organization across all free models. |
|---|
| Evidence note | documented limits and the no-card statement are quoted from the vendor's docs and pricing pages; no call was made. |
|---|
| Parsed numbers | - request-rate 200 requests / day
|
|---|
Evidence 6
Unknown values 3
$.egress.metered egress metering was not published$.per_object_cap.stated the field does not apply to this kind of offer$.retention.stated the field does not apply to this kind of offer