Grok 4.7 Fast is only in Cursor and Grok Build, at 2x the rate
xAI's pricing page says Grok 4.7 Fast is not on the public API. Where it runs, what it costs, and what to do if your Sume agent needs a Grok model.
Grok 4.7 Fast is not something you can call on xAI's API. xAI's pricing page lists it as available only in Cursor and Grok Build, at twice the Grok 4.7 rates, and 1.5 times for long-context prompts (read 2026-10-04). The model you can call with grok-4.7 is the standard one, with a 500k context window.
If you are wiring an agent that makes video, the practical question is which of these can be the orchestrator of a Sume run, and the answer comes from Sume's docs, not xAI's.
What does the standard Grok 4.7 cost?
From xAI's pages, read today:
| Item | Price |
|---|---|
| Input | $2 |
| Output | $6 |
| Input and output over 200k prompt tokens | $4 and $12 |
| Cached input | $0.50 |
| US regional endpoint | 1.1x |
| Fast variant | 2x, or 1.5x long-context; Cursor and Grok Build only |
Can a Sume Format run use Grok?
Only if the id is in the Agents catalog. Sume's Create a run page says model takes a catalog id and anything else is a 400 invalid_request. Sume's docs do not list grok-4.7 as a Format run model, so do not assume your workspace can select it; read the response to a small probe run, and cancel it if accepted. No Fast id is documented.
What if I want Grok's speed anyway?
Use Cursor or Grok Build for the coding or planning step and send the result to Sume as input. Sume treats input as data, written whole to the run's workspace, so a plan produced elsewhere is safe to hand over without being mistaken for instructions.
Sources
Related posts
More in Agents
- Hermes cron fixes for run now, and Sume's one-active-run rule
Hermes fixed run-now cancelling the next scheduled run. Sume schedules allow one active run: a second is skipped, or rejected if you ask for reject.
- job_scope_forbidden: why an agent run cannot list every Sume job
A thread-bound Sume credential lists only its own thread's jobs, and unattended runs cannot ask for the whole workspace. What to use instead.
- Kilo scheduled sessions vs a Sume schedule for recurring renders
Kilo v7.8.3 reports a session waiting on a wakeup or cron task as scheduled. A Sume schedule lives server-side; use it when renders must run unattended.
- OpenAI Dots x Runway: brief a dot, it plans shots. The Sume loop
Runway's Oct 1 changelog lists OpenAI Dots x Runway for all plans: brief a dot, it plans shots, Runway shoots. The same loop with Sume MCP tools.
Written by Sume