Mistral Large 4 preview: Sume's agent model field takes one value
Mistral Large 4 is reported in preview. Sume lists no Mistral Large id; Agent Completions accepts only sume-agent. What to send instead, and what it bills.
The AI Brief of October 7, 2026 reports a Mistral Large 4 preview. If you searched for it next to the word Sume, the short answer is: there is no Mistral Large entry in the Sume repo as of 2026-10-07, and you cannot pick one.
The Agent Completions docs say the model field accepts only sume-agent, and that omitting it gives the same agent. A request with any other model returns 400 invalid_request. So there is no model dropdown to wait for a Mistral row in.
What to do if you want Mistral in the loop
Keep Mistral as your own planner and let it call Sume for the media work. Two existing paths are covered in the posts on Mistral connectors and Mistral Vibe: both treat Sume as a tool server, with the paid tools behind a confirmation or a read-only allowlist.
- Planner: Mistral (your account, your bill, outside Sume).
- Media jobs: a Sume Format run or a direct model call, billed to the Sume USD wallet.
- Cap: every Agent Completion needs
generation_spend_cap_usd; there is no default.
Two meters, not one
Whatever planner you use, Sume only meters generation. The table lists what a run can spend on, using the catalog rates read 2026-10-07.
| Job | Billable rate |
|---|---|
| Wan 3.0 720p | $0.125 per second |
| Kling 3 with audio | $0.21 per second |
| Timeline render | $0.10 per output minute |
| Text to speech | $0.0475 per 1,000 characters |
Sources
Related posts
More in Agents
- One webhook handler for Sume agent, Format and schedule runs
A single receiver can handle all three Sume run types: verify the HMAC over timestamp.body, dedupe on request_id, and branch on outcome (ok, degraded, error).
- Polling a Sume agent run: next_action values and backoff
Poll Sume runs by branching on next_action (poll_status, retry_later, none), using status_url and result_url, backing off on 429 and 503.
- Scheduled or Format API: does the clock or your user start the run?
A Sume Scheduled run fires on a cron; a Format run fires when your backend calls it. Same agent, same receipt, new trigger. How to choose without dupes.
- script_run or bulk runs: where to fan out 20 ad variants
script_run fans out three or more Sume tool calls inside one 55-second request; bulk runs queue up to 100 Format runs for hours. Limits and a rule of thumb.
Written by Sume