Higgsfield MCP has no credit cap: cap spend with Sume max_spend_usd
Higgsfield's MCP guide says there is no built-in credit spending cap. Sume's MCP tools take dry_run and max_spend_usd. How they work, and what they skip.

Higgsfield's MCP-for-marketers page says the connector has no built-in credit spending cap, and that an active subscription is required. If you let an agent make video, that matters. Sume's MCP generation tools take dry_run=true, which previews admission or cost without submitting, and max_spend_usd, which is enforced only when you provide it.
The facts
Quoted from the Higgsfield page and the Sume MCP tools and gates docs.
| Fact | Value |
|---|---|
| Higgsfield MCP | No built-in credit spending cap |
| Higgsfield MCP access | Active subscription required |
Sume dry_run | Preview cost or admission, no submit |
Sume max_spend_usd | Enforced only when provided |
Sume idempotency_key | Available on generation tools |
A safe agent loop
First call the tool with dry_run=true and read the cost. Then call it again with the same inputs, an idempotency_key and max_spend_usd set to the estimate plus a small margin. The agent cannot spend more than that on that call.
Note the wording: the cap applies only if you pass it. An agent prompt that forgets the field has no cap on that call, so put the rule in the tool wrapper rather than relying on a reminder.
What the cap does not do
max_spend_usd bounds one request. It is not an account budget. For a campaign limit, count spend in your own code across calls, and rely on plan concurrency and queue limits for volume.
Avatar tools in the docs include avatars_create, avatar-videos_create and avatar-video-previews_create.
Sources
Related posts
More in Agents
- hypit Understand order: one probe, then parallel batches
Sume's hypit Understand order: probe alone, then transcribe, boundaries and tiles in one batch, then notes. Which verbs wait on the transcript.
- How an agent picks a video model over hosted MCP
An agent reads video-router_models, picks an id, then calls generate_video with an idempotency_key and waits with jobs_wait. Omit the model for sume/auto.
- Get a typed podcast clip plan from an Agent Completion
Send a transcript and an output_schema to POST /v1/agent/completions, and get back start and end times for each clip, ready for video-trim.
- Poll a Sume Action run in Python: branch on next_action, not status
A small Python loop for a Sume Scheduled run that sleeps on poll_status, retries on retry_later and stops on none, then fetches the result.
Written by Sume