output_schema_invalid 400: read violations path and rule tokens
A Sume Format run with a schema outside the supported subset returns 400 output_schema_invalid with every violation listed. Switch on rule, never on message.
Formats
Sume Formats over the API: create and version a Format, call it, run bulk batches, get structured output, take webhooks, and debug failed runs.
Start with: What is a Sume Format? Turn an agent thread into one API call
Sume output_schema limits: depth 10, 5000 properties, 120,000 chars
Sume's output_schema caps nesting at 10 levels, 5000 properties, 1000 enum values and 120,000 characters of strings. See which rule fires and how to shrink.
Brand setup for AI ad batches: a saved Format as the brand kit
Higgsfield Ads Studio starts with brand setup. On Sume, a saved Format holds your house style and output contract so each ad run starts from it.
White label AI video generator: build it on an API
Sume documents no white-label program, but its API lets you run AI video for your clients under your brand: one server key, per-run caps, files you host.
AI luxury product ad: a cinematic studio commercial by API
A luxury-style product ad can be one API call. Sume's sume-cinematic-studio-commercial Format makes a studio commercial from a packshot; the call and limits.
AI UGC video generator API: one call to a close-camera Format
A UGC-style video generator can be one API call. Sume's sume-close-camera-ugc Format makes a handheld selfie-style video; see the request and the receipt.
403 insufficient_scope on a Sume Format run: scope or service key
A 403 insufficient_scope on a Sume Format call means the key lacks formats:write or is a service-account key. How to tell which, and the fix: a new key.
How long can a Sume Format run take? The expires_at deadline
A Sume Format run is force-finalized as failed 90 minutes after it was created, or sooner if it goes silent. How to read expires_at and set your own timeout.
403 workspace_key_required: call a team Format with a team key
A 403 workspace_key_required means a personal key called a team workspace's Format. Create a key inside that workspace; details.workspace_id names it.
Which model runs a Sume Format? The model field on a run
The model field on a Sume Format run picks the LLM that orchestrates it, default gpt-6-sol. It does not pick the image, video or audio models. Rules and errors.
JSON schema for AI video output: 5 templates to copy
Copyable JSON Schemas for AI video output on Sume: one video, video plus copy, a poster, aspect ratios, and caption cues, all in the strict subset.
Regenerate one scene of an AI video with the Sume API
To regenerate one scene of an AI video, continue the Sume Format run with previous_run_id and name the scene. It can redo that clip and leave the rest alone.
Sume Format run failed: error codes, details, and what to do
A failed Sume Format run names its cause in error.code. Here is each code with its details fields, what gets billed, and whether to continue or retry.
Prompt injection and user input: pass data to an AI agent API
Put customer text in a Sume run's input object, not its instruction. The agent is told to read it as data; the docs call that a trust boundary, not a sandbox.
How to write a SKILL.md file for a Sume Format recipe
A Sume Format recipe is a SKILL.md file plus references. Keep SKILL.md a short index, move detail into references/, and follow the package rules.
How to verify an AI-generated video was delivered by a Sume run
Before you count a Sume run as delivered, check status and primary_output_url, read filled_by, and probe the file. Here is what Sume checks, and what it skips.
Get partial results from a failed AI video run with a scene ledger
A failed Sume Format run still reports the scenes it made if your output schema allows a partial. Use nullable fields, skip minItems, and name a primary key.
Optimistic concurrency with If-Match: edit Sume Format files by API
Edit a Sume Format's files over the API without overwriting another writer: each write carries the file's blob sha, and If-Match guards the whole package.
Create an AI video workflow via API: a Sume Format and its SKILL.md
Create a Sume Format with POST /v1/formats, then write its SKILL.md recipe through the Contents API. The fields, the slug rules, and the errors to handle.
Share an AI video workflow with another team: Sume Format grants
Share a Sume team Format with another team workspace. Once the grant is live, that team calls the owner's address with its own team key and pays for its runs.
Sume API limits for Formats: caps, sizes, and timeouts
The documented Sume Format API limits in one place: request and input sizes, attachments, schemas, spend caps, queues, run deadlines, webhooks, rate limits.
JSON Schema oneOf not supported in strict mode: fix the violations
Sume rejects an output_schema outside the OpenAI strict-mode subset with 400 output_schema_invalid. Read each violation's rule, then port oneOf and nullable.
List Formats API: find and address Sume Formats by handle or id
GET /v1/formats lists the Formats your key can call. Address one by {handle}/{slug} or by its permanent skl_ id, and list its runs newest first.
Why did my AI agent ignore my instructions? Debug a Sume Format run
A checklist for a Sume Format run that did something unexpected: the composed first message, the carried instruction, the version, events, and output.
Version control for AI agent prompts in production: Sume Formats
Every Sume Format edit bumps its version, and each run receipt records the version that ran. Runs cannot be pinned to a version, and there is no rollback API.
Retry failed items in an AI video batch: Sume bulk-run recovery
When a Sume bulk queue completes, read counts.failed, then continue each failed child that left work behind or resubmit it under a new Idempotency-Key.
Which tool to generate UGC-style video ads for products at scale?
It depends on your input, presenter, languages, and volume. Arcads, Creatify, HeyGen, Runway, Synthesia, and Sume compared on API, pricing, and limits.
Sume Format structured output: typed JSON from a JSON Schema
Bind a JSON Schema as output_schema and a Sume Format run returns output in that shape, with each media URL checked against what the run made.
Sume Format bulk runs: queue up to 100 renders in one request
A Sume bulk request queues 1 to 100 ordinary Format runs on the server and keeps 1 to 16 in flight. Poll one queue URL; read each child as a normal run.
Sume Format run lifecycle: statuses, webhooks, and polling
A Sume Format run goes from queued to processing to a terminal status. Learn the outcome from one signed format.run.terminal webhook, by polling, or both.
How to embed AI video generation in your product with Sume Formats
To embed AI video generation, your server holds one Sume API key and runs a Format per customer, with a derived Idempotency-Key, spend cap, and webhook.
What is a Sume Format? Turn an agent thread into one API call
A Sume Format is a saved video recipe your backend calls by handle and slug. One POST runs it in a fresh sandbox and returns media plus optional typed JSON.
Ready-made Formats for product video: the Sume Format catalog
Sume ships ready-made Formats for product and UGC-style video and images, each callable from your backend with one HTTP request at the reserved sume handle.