# Fix list: Getty Images MCP Server From Anchor Terminal's listing at https://www.anchorterminal.com/tools/getty-images-mcp, the October 2026 research run, assessed 27 October 2026. Grade F, 32.4 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 Getty Images MCP Server: 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. Schema & documentation, 19 out of 100, up to 13.2 more on the total Why it scored 19: No machine-readable contract for the MCP tools could be read. `tools/list` answers 401 without credentials and no tool list, server card or source is published. The REST API the catalogue also sits behind has a public OpenAPI 3.0.4 file of 66 paths covering search, metadata and downloads, but how the tools map to it isn't documented. Judgement call, 8 of 25. No llms.txt (404 on both hosts) and no Markdown docs (0). The developer page names four jobs (authentication, search, metadata, download) with no per-tool descriptions or guidance on when not to use them (3 of 20). Typed inputs unread (0 of 15). No MCP examples; the API docs show the 401 and 429 responses (3 of 15). The endpoint carries /v1 and the API has dated release notes, with no MCP entry (5 of 15). Total 19. 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. ## 2. Payments & pricing, 0 out of 100, up to 12.5 more on the total Why it scored 0: No x402, MPP or L402 (0). No price is published for the MCP server or for the Premium Access and Editorial Subscription agreements it needs; the route in is a contact-sales form (0). No free tier or self-serve trial; the API page says a test account is available on request (0). A person has to hold a Getty agreement and get credentials from an account representative (0). Total 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. ## 3. Agent ergonomics, 24 out of 100, up to 12.4 more on the total Why it scored 24: Graded on the MCP server without reading its tools. The tool count isn't published; the developer page names four jobs, which suggests a small set, but neither count nor size was checked, so 10 of 25. The REST API has field selection (summary_set, detail_set, named fields), paging and many search filters, but whether the tools expose them is unchecked (5 of 20). The MCP endpoint's 401 carries no WWW-Authenticate header, and the API's documented errors are the 401 body and the 429 message (4 of 20). No idempotency guidance and annotations unread; downloads draw on the customer's agreement, so a repeated download isn't free of consequence (0 of 20). Official SDKs and samples exist for the REST API, none for the MCP server, and a token must be fetched with a key and secret first (5 of 15). Total 24. 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. ## 4. Security & auth, 40 out of 100, up to 10.5 more on the total Why it scored 40: OAuth 2.0 with an API key and secret per customer, tied to the agreement and issued by Getty. The metadata lists the authorisation code grant with PKCE S256, client credentials and refresh tokens, with tokens of at most 30 minutes per the API docs and refresh tokens Getty or the user can revoke. No scopes, dynamic client registration or revocation endpoint are published, and the client credentials token reaches the whole agreement. Secrets go in the body or a Basic header, not a URL (20 of 30). The jobs are search, metadata and download; no read-only mode or confirmation for downloads was found (5 of 20). Results carry captions, titles and keywords written by contributors, and Getty says AI may help draft captions with human review. No prompt-injection guidance (5 of 15). The API exposes download and purchase history per account, which gives some per-call visibility; no MCP audit log found (5 of 15). security.txt is valid with a contact and expiry only; no bug bounty, certification or public advisories found (5 of 20). Total 40. 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. Reliability, 50 out of 100, up to 10 more on the total Why it scored 50: Graded with the hosted lines, on the remote MCP server. Getty keeps a dated status and notifications page on the developer site for the API and websites, with posts back to 2023 but no components and no mention of the MCP server. Judgement call, 10 of 20 for a page that doesn't name the surface graded, as for Functionize. From 29 July to 27 October 2026 it records one incident, 500 errors across the API and websites from about 05.45 to 08.20 Pacific time on 27 August, which is a major outage by the checklist (10). The developer page gives 20 queries a second for the MCP server (15). The API docs show the 429 response and its error detail for the per-second limit, with no Retry-After, backoff or idempotency guidance; downloads are POST calls (5 of 15). No SLA found (0). The release of 12 August 2026 calls it launched and nothing labels it beta, though the developer page says tools will change in the coming months (10). Total 50. 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. ## 6. Maintenance & community, 28 out of 100, up to 6.3 more on the total Why it scored 28: Read as a closed hosted service. The MCP server launched on 12 August 2026, 76 days ago, and the developer page's sitemap date is the same day (20). No dated MCP change since the launch; the API release notes' newest entry is 2 March 2026 (0 of 20). The API has a public release notes page, Getty says sales and technical support teams work around the clock, and MCP feedback goes through account representatives (5 of 15). Not in the official MCP registry (searches for getty and gettyimages returned nothing on 27 October 2026) (0). No package for the server; Getty's public sample-code repository for the REST API was updated to .NET 10 in April 2026 and last changed on 30 September 2026 (3 of 10). Total 28. 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. ## 7. Transparency & trust, 68 out of 100, up to 2.8 more on the total Made of editorial 35, provenance 100. Why it scored 68: The editorial half. A closed service. The website terms (August 2025) and the content licence agreement (April 2026) are public and state the AI limits plainly, but the Web Service Terms that govern MCP access aren't published (10 of 30). The privacy policy (March 2026) states general retention, primary data centres in the United States, the US affiliates data may go to, Data Privacy Framework adherence and Standard Contractual Clauses. Nothing says what the MCP server logs from queries, and the website terms give Getty a licence to use search terms for any business purpose on the site (12 of 30). No deprecation policy. The API release notes give some dated notices, such as the order_id change of 2 March 2026, and the developer page warns tools will change without saying how (5 of 20). Data locations stated in general terms and affiliates named, with no subprocessor list (8 of 20). Total 35. 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. ## 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: the tools as served, because `initialize` and `tools/list` answered 401 without credentials and no tool list, schema or source is published. Names, count, descriptions, inputs and annotations are unread, and so is the transport. - unchecked: the Web Service Terms, which the MCP FAQ says must be accepted before access. No public copy was found on gettyimages.com, the developer site or by search. - unchecked: whether the MCP server also needs the `Api-Key` header the REST API requires, or only a bearer token. The developer page says the agent finds authentication details at the endpoint, which answers 401 without a body to read. - unchecked: whether customers on self-serve Getty subscriptions, rather than Premium Access or an Editorial Subscription, can get MCP credentials. The pages name an agreement and an account representative. - Listed rather than held. Access runs through a Getty agreement and an account representative, but the server is generally available to existing customers and partners, with a public developer page, a live endpoint and public OAuth metadata. That is more than an announcement, by the rule for limited rollouts to customers of a product companies can buy. - firstReleased is 12 August 2026, the release date, matching the developer page's sitemap date (12 August 2026, 08.20 Pacific). That agrees with the lead. - Category. data-providers, the closest fit, because Getty licenses a catalogue of content and metadata behind an API and the data-quality fields record coverage, licence and reuse, which matter most here. No category or capability covers stock images and video; capability data.knowledge, as for Wikimedia, is a poor fit. search was the alternative and covers web search. Judgement call. - Getty's claims that its library doesn't allow AI-generated content and that the server isn't a generative AI service are recorded as claims. Getty's own AI generator output is licensed separately and wasn't checked against the library. - Getty's website terms forbid data mining and robots on the site. Recorded as a fact; only public pages that robots.txt allows were read, one request each. - Our own products. The Getty Images MCP Server doesn't do LocalGhost's job or letme's. It reaches one vendor's licensed catalogue and chooses no tools. Graded the normal way. ## Weaknesses - Access needs a Getty agreement such as Premium Access and credentials from an account representative; no self-serve sign-up, trial or published price - Tool names, schemas and annotations aren't public; `initialize` and `tools/list` answer 401 without credentials - The Web Service Terms a customer must accept before access aren't published - No scopes, dynamic client registration or revocation endpoint in the authorisation server metadata - API and websites returned 500 errors for about 2 hours 35 minutes on 27 August 2026, per Getty's status page ## 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. - Ask the person for the Getty API key and secret, which only their Getty account representative can issue after the Web Service Terms are accepted - Exchange the key and secret for a bearer token at authentication.gettyimages.com and reuse it until it expires, since token requests count against the rate limit - Read `tools/list` at the start of each session rather than hard-coding tool names, because Getty says tools and parameters will change - Check whether an asset is editorial before suggesting commercial use, since editorial content isn't licensed for advertising or promotion - Don't pass downloaded images or their metadata to a model or tool that trains on inputs; the licence bars training and ends rights if an AI tool uses content for itself ## 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.