OpenAI Image API by OpenAI

# Fix list: OpenAI Image API

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/openai-image-api, the October 2026 research run, assessed 1 October 2026. Grade BB, 72.7 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 OpenAI Image 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-token prices published without a login. No per-image figure for GPT Image 2.5, but the unit price is public (20). No free tier, prepaid credit needed (0). Human sign-up and organisation verification before the first GPT Image call (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. Agent ergonomics, 72 out of 100, up to 4.6 more on the total

Why it scored 72: Output size is controlled with `size`, `quality`, `output_format`, `output_compression` and `n`, but images come back as base64 only, so a raw response is large (18). Output-size controls, nothing to paginate (15). Error types an agent can branch on, with moderation separated from server faults (16). No idempotency key on image endpoints, so a retried generation that had succeeded is billed again. Backoff guidance is in the rate-limit guide (8). Only `model` and `prompt` required, official Python and TypeScript SDKs among others (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.

## 3. Reliability, 80 out of 100, up to 4 more on the total

Why it scored 80: Public status page at status.openai.com with incidents filed per component (20). Three API image incidents in the last 90 days, elevated errors on 21 July (error rates back to normal after about 53 minutes, closed after 2 h 46 min), gpt-image-2 errors for 13 minutes on 24 July and gpt-image-2.5-flare errors for 40 minutes on 16 September. ChatGPT had four more image incidents that didn't touch the API. Minor only (20). Image limits published per tier, 5 to 250 images a minute on GPT Image 2.5 Flare (15). 429 handling documented with Retry-After, x-ratelimit headers and backoff code (15). The 99.9 per cent Scale Tier SLA doesn't cover image models (0). GA (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.

## 4. Security & auth, 85 out of 100, up to 2.6 more on the total

Why it scored 85: Project API keys with restricted per-endpoint permissions, service-account keys and revocation (30). Read-only key permissions, per-project model allow-lists and spend limits. No destructive actions on the image endpoints (15). Returns generated images only, no third-party content (10). Audit Logs API records key and project events, and usage is broken down per project and key, but there's no per-request log (10). security.txt valid, a public bug bounty, SOC 2 Type 2 and a trust portal (20).

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.

## 5. Maintenance & community, 79 out of 100, up to 1.8 more on the total

Why it scored 79: GPT Image 2.5 Sunburst and Flare released on 8 September 2026, 23 days ago (30). Model churn in place of release counts. Notice for GA image models meets OpenAI's six-month policy (gpt-image-1 announced 22 April for 23 October, three more announced 2 June for 1 December), but six image models leave the API within seven months (12). Public changelog, developer forum and support (12). openai-node shipped nine releases between 26 August and 10 September, npm is at 7.25.0 (15). SDK packages current with CI (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.

## 6. Schema & documentation, 90 out of 100, up to 1.6 more on the total

Why it scored 90: OpenAPI spec published in openai/openai-openapi (25). llms.txt at developers.openai.com (10). The guide says when to use Sunburst (precise edits) and when Flare (fast everyday generation), and when to use the Image API versus the Responses tool (15). Typed parameters with enums for size, quality, output_format, background and moderation, though several default to `auto` (13). Examples in every language and named error types such as `moderation_blocked` and `image_generation_user_error`, but no full error catalogue for images (12). Dated snapshots and a public changelog (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.

## 7. Transparency & trust, 92 out of 100, up to 0.7 more on the total

Made of editorial 83, provenance 100.

Why it scored 92: Closed models, clear Services Agreement that assigns output to the customer (15). The data controls page gives per-endpoint retention, 30-day abuse logs, no training on API data by default and ZDR eligibility for every GPT Image model, consistent with the policies (28). Deprecation policy with notice periods by model class and a dated table (20). A subprocessor list of 18 vendors with processing locations, updated 9 July 2026, and data residency in ten regions for image endpoints (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.

## Deductions

Each comes off the total. A fixed and documented problem counts for less at the next check.

- On 2025-11-26 OpenAI disclosed that its analytics vendor Mixpanel was breached, exposing names, emails, rough location and IDs of API platform users. No API keys, prompts or outputs. Mixpanel was removed and the incident documented (https://openai.com/index/mixpanel-incident/). -2 because it touched every API account, decayed because it's fixed and disclosed.

## 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.

- No published statement on whether a call blocked by moderation is billed.
- The listing's lastRelease of 2026-09-22 couldn't be matched to an image change. The last image entry in the changelog is 2026-09-08.

## Weaknesses

- No per-image price for GPT Image 2.5, and the cost calculator doesn't cover it
- Tier 1 accounts get 5 images a minute on GPT Image 2.5 Flare
- Base64-only responses put the whole image in the payload
- Six image models removed or scheduled for removal between 2026-05-12 and 2026-12-01
- Three API image incidents in the last 90 days, the longest about 53 minutes of elevated errors

## 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.

- Pin `quality` and `size`. `auto` makes cost and latency vary per call
- Read `usage` in the response to track image tokens, it's the only cost signal for GPT Image 2.5
- Don't retry `moderation_blocked` or `image_generation_user_error`. Change the prompt first
- Honour `Retry-After` on 429 and add jitter, image limits are per minute and low at tier 1
- Move off gpt-image-1 before 2026-10-23 and off 1-mini, 1.5 and chatgpt-image-latest before 2026-12-01

## What the review panel asked for

- URL response option
- Idempotency key
- add 2.5 to the cost calculator
- state whether moderated calls bill

## 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.