Higgsfield API

# Fix list: Higgsfield API

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/higgsfield, the October 2026 research run, assessed 8 October 2026. Grade B, 63.5 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on Higgsfield API: work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Payments & pricing, 20 out of 100, up to 10 more on the total

Why it scored 20: No x402, MPP or L402 (0). Per-second and per-image list prices for each model are public at open.higgsfield.ai/pricing without a login (20). No free credits or card-free trial were found. A new account gets sign-up discounts and 2 concurrent generations, and usage draws on a prepaid balance (0). A person signs up in a browser and creates the key in the console (0).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):

The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).

- 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which.
- 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login.
- 20, a free tier or trial that doesn't need a card.
- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).

Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied.

Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol.

## 2. Security & auth, 48 out of 100, up to 9.1 more on the total

Why it scored 48: Hosted reading. A key ID and secret pair created and managed in the console, scoped to a workspace, with rotation advised and no permission scopes. Secrets travel in a header, not the URL (20). No read-only or restricted keys and no confirmation step. Organisations have admin and member roles, and an estimate call shows cost before spending (3 of 20). The API returns generated media and no third-party text (10). Request history and usage analytics in the console and an `X-Correlation-ID` on every response, no audit log found (8 of 15). A valid security.txt with a PGP key, expiring 12 September 2027, and a disclosure clause in the terms. No bug bounty found. The trust centre at trust.higgsfield.ai is drawn by script and was unread, so certifications are unchecked (7 of 20). No webhook signature is documented.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-security):

- 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option.
- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.
- 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10.
- 0 to 15, audit logs or per-call visibility for the operator.
- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.

Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing.

## 3. Schema & documentation, 69 out of 100, up to 5 more on the total

Why it scored 69: `/docs/openapi.json` is public OpenAPI 3.1 but covers 8 paths, and llms.txt calls it supplementary. Each model page carries its own full JSON Schema (15 of 25). llms.txt, llms-full.txt and Markdown twins of every page (10). Model pages state the endpoint, usage notes and limits, with little on when to pick one model over another (14 of 20). Schemas have enums, length and item limits, required fields and `additionalProperties: false` (14 of 15). cURL, Python and TypeScript examples on each page and a status-code table, but errors are a FastAPI `detail` string with no machine codes (11 of 15). Endpoints carry a version segment, and no changelog was found in the docs or sitemap (5 of 15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):

APIs and MCP servers.

- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).
- 10, llms.txt or Markdown docs served for agents.
- 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference.
- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.
- 0 to 15, examples and documented error responses.
- 15, versioning and a public changelog.

Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference.

## 4. Agent ergonomics, 77 out of 100, up to 3.7 more on the total

Why it scored 77: Submissions return a small JSON handle, results are URLs, and webhooks replace polling (20 of 25). Resolution, duration and aspect ratio set output size per model. No endpoint to list requests was found in the docs (10 of 20). The error page says which statuses to retry, but the concurrency limit arrives as a 400 with a message string and the docs warn against parsing messages (12 of 20). `Idempotency-Key` on every generation endpoint, with a 422 on a reused key with different parameters, plus cancel for queued requests (20). Few required parameters, sensible defaults, and official Python and TypeScript SDKs (15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):

- 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries).
- 20, pagination, filtering and output-size controls.
- 20, actionable, documented error responses, codes and messages an agent can recover from.
- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.
- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.

Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs.

## 5. Transparency & trust, 61 out of 100, up to 3.4 more on the total

Made of editorial 45, provenance 77.

Why it scored 61: Closed service with API terms dated 2 September 2026 that are specific on licence, resale and outputs. SDKs are MIT and Apache-2.0 (15). The API terms state training on inputs and outputs by default with a workspace opt-out, request records kept while the account is active, media kept at least seven days and zero retention only by enterprise addendum. The privacy policy agrees on training. The DPA the terms incorporate was not found (20 of 30). The terms promise commercially reasonable efforts to announce model removals through a changelog, no changelog or dated notice was found, and legacy endpoints are labelled in the docs (5 of 20). Partner models are named by vendor in the catalogue. The subprocessor list is said to sit in the trust centre, which was unread, and no data locations were found (5 of 20).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):

- 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms.
- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).
- 0 to 20, a deprecation policy or notices with dates.
- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).

The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.

Provenance checks not met in full (half of this category, computed from checked facts):

- Domain age: higgsfield.ai, registered 2023-06-07 (3 years) (7 of 15)
- Terms of service: read, states 6 of the 7 things a reader expects, and has 1 clause that costs points (7.1 of 10)
- Privacy policy: read, states 8 of the 8 things a reader expects, and has 1 clause that costs points (8 of 10)
- Changelog: not found (0 of 10)

## 6. Reliability, 85 out of 100, up to 3 more on the total

Why it scored 85: Hosted reading. status.higgsfield.ai on Instatus with API and Cloud Console components and a notice history (20). No notices from August to October 2026 and 100.0 per cent shown for both components (30). Concurrency limits are published with numbers, 2 for a new account and 10, 20 and 40 at $25, $100 and $1,000 of funds added in 28 days (15). The docs give backoff with jitter, a retry table and idempotency keys for submissions, but reaching the limit returns 400 and no `Retry-After` or rate-limit headers are sent (10 of 15). The API terms say self-serve customers get no uptime commitment and an SLA exists only under an order form, with none published (0). No beta or preview label on the API (10).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):

Hosted APIs, MCP servers, models and platforms.

- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).
- 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so.
- 15, rate limits documented with numbers.
- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.
- 10, an SLA published for any paid tier.
- 10, the surface agents use is generally available, not beta or preview.

Local packages, SDKs, frameworks and stdio MCP servers.

- 20, installs from an official package with supported runtimes stated.
- 25, a public CI and test suite, passing on the default branch.
- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).
- 15, semver discipline and breaking changes called out in a changelog.
- 15, version 1.0 or later, or declared stable.

Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.

## 7. Maintenance & community, 75 out of 100, up to 2.2 more on the total

Why it scored 75: The docs sitemap shows pages changed on 7 October 2026, and the SDKs were last released on 17 September 2026 (30). `@higgsfield/client` 0.2.4 and 0.2.6 and `higgsfield-client` 0.2.0 all shipped in September 2026 (20). No public changelog. Support is by Discord and email, and the SDK repositories have 5 and 14 open issues whose replies we couldn't read (6 of 15 for a closed service). Official Python and TypeScript SDKs are current (15). The TypeScript repository has a publish workflow and the Python one has tests but no CI workflow (4 of 10).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):

- 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older.
- 20, at least three releases or dated changelog entries in the last 90 days.
- 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15.
- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).
- 10, package health, current dependencies and CI.

Models are read for deprecation notice periods and model churn rather than release counts.

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- unchecked: trust.higgsfield.ai is drawn by script, so certifications, the subprocessor list and security practices were unread
- unchecked: the FAQ answers on open.higgsfield.ai/pricing (launch discount, agent use, spending controls) load by script, so the conditions for the max-discount prices weren't read
- unchecked: the DPA that the API terms incorporate. No public link to it was found
- unchecked: reply times on the SDK repositories' open issues (5 on higgsfield-client, 14 on higgsfield-js)
- Whether a new account receives any free credits, and whether a card is needed before a first call. Neither was stated on the pages read
- The API terms say requests over the concurrency limit are queued or rejected with a retryable error, and the docs say the API returns 400
- The API terms mention a changelog, test keys that run in dry-run mode and a `/v1/` version prefix. None of the three was found in the docs
- The lead named the vendor Higgsfield AI. The legal entity in the terms is Higgsfield Inc. The rest of the lead held

## Weaknesses

- API inputs and outputs can train Higgsfield's models by default. The opt-out is a workspace setting, processed within 10 business days
- Reaching the concurrency limit returns `400 Bad Request` with a message string, and the docs say no rate-limit headers or `Retry-After` are sent
- No changelog was found in the docs, although the API terms name one as the channel for deprecations and removals
- Keys have no scopes or read-only mode, and no webhook signature is documented
- No free credits were found, and a new account starts at 2 concurrent generations until it adds $25
- `/docs/openapi.json` covers 8 paths of an 84-entry catalogue, and llms.txt calls it supplementary

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Send `Authorization: Key <key_id>:<key_secret>` to https://api.higgsfield.ai. A Bearer token is rejected
- Send a fresh `Idempotency-Key` with every generation. After a timeout or 5xx, repeat the same request with the same key, then read `status_url`
- Treat a 400 whose `detail` mentions concurrent requests as a limit, not a bad request. Wait for a running job to finish before submitting more
- Poll `status_url` from 2 seconds up to 10 with jitter, and stop on `completed`, `failed`, `nsfw` or `canceled`
- Download output within seven days. Media files may be deleted after that
- Read the model's own docs page for its schema. The OpenAPI file lists only a few models

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.