Rate Limits
Talonic meters usage in credits, not request quotas. Every account starts with 5,000 free credits per month, then you pay as you go with prepaid credits. Separately, each workspace has daily request limits per endpoint family. They exist to keep the platform healthy and are not pricing.
Credit metering
Every operation costs credits (1,000 credits = €1): 200 credits per page ingested, 20 credits per AI-resolved cell, and registry-resolved cells are free. See pricing for the full rate card. When the balance cannot cover a request, the API returns 402 with code INSUFFICIENT_CREDITS until your free credits reset or you buy credits inside the platform.
Daily request limits
Limits count requests per workspace per UTC day, separately for each endpoint family, and reset at midnight UTC. They depend on your API tier (throughput tier, not your billing plan):
| Endpoint family | Free | Pro | Enterprise |
|---|---|---|---|
| Extract (POST /v1/extract) | 1,000 | 25,000 | Unlimited |
| Run (POST /v1/run) | 1,000 | 25,000 | Unlimited |
| Ask (POST /v1/ask) | 200 | 5,000 | Unlimited |
| Ask streams (GET /v1/ask/:id/stream) | 400 | 10,000 | Unlimited |
| Upload / ingest / operations | 1,000 each | 25,000 each | Unlimited |
| Agent tool invocations | 500 | 20,000 | Unlimited |
| Other platform endpoints | 2,000 | 50,000 | Unlimited |
File size limits
Uploads are capped per API tier: 20 MB on Free, 100 MB on Pro and 500 MB on Enterprise. 500 MB is the hard maximum for any single file on /v1/extract, /v1/run, /v1/process and source uploads. A file over your tier's cap is refused before any work with 413 and code FILE_TOO_LARGE. POST /v1/run accepts up to 20 files per call.
Rate limit headers
Rate-limited responses carry headers so you can pace bulk extraction workloads before you hit the limit.
| Header | Description |
|---|---|
| X-RateLimit-Limit | Daily limit for this endpoint family on your API tier |
| X-RateLimit-Remaining | Requests left today |
| X-RateLimit-Reset | ISO 8601 timestamp of the next reset (midnight UTC) |
Handling 429 responses
A 429 Too Many Requests response means today's limit for that endpoint family is used up. The error body's details carries limit, used and reset_at: wait until reset_at rather than retrying in a loop. See the error handling guide for the full retry strategy.
Best practices
Send many documents in one POST /v1/run call (up to 20 files) instead of many single calls. Use webhooks to receive results asynchronously rather than polling jobs. Cache schema responses to avoid unnecessary reads.