# Fix list: Google Lyria From Anchor Terminal's listing at https://www.anchorterminal.com/tools/google-lyria, the October 2026 research run, assessed 1 October 2026. Grade D, 49 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 Google Lyria: 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. Reliability, 33 out of 100, up to 13.4 more on the total Why it scored 33: AI Studio has a status page, but it renders only in a browser and we couldn't read components or history from it (10). No readable incident history (5). Rate limits are defined per project and usage tier as RPM and RPD, but the docs give no numbers for Lyria and point to a dashboard behind sign-in (5). A 429 `RESOURCE_EXHAUSTED` is documented and the troubleshooting guide asks for exponential backoff with jitter (8), with no idempotency or safe-retry guidance (0). No SLA found for the Gemini API (0). `lyria-3.5` is GA since 2026-09-03, but it's served from the `v1beta` Interactions endpoint and the clip model is still a preview (5). 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. ## 2. Payments & pricing, 20 out of 100, up to 10 more on the total Why it scored 20: No x402, MPP or L402 (0). Per-song and per-clip prices published without a login, $0.08 a song on `lyria-3.5` and `lyria-3-pro-preview` and $0.04 a clip on `lyria-3-clip-preview` (20). Lyria has no free tier, so the key's Cloud project needs billing (0). A person creates the key in AI Studio and attaches billing (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, 40 out of 100, up to 9.8 more on the total Why it scored 40: Audio comes back as base64 inside the JSON response, which an agent has to decode and keep out of its context (5). `response_format` picks WAV over the default MP3 on `lyria-3.5`, but length can't be set except in the prompt (5). No list, pagination, webhook or progress controls for music (0). Errors follow Google's standard code, status and message shape, with `RESOURCE_EXHAUSTED` for quota and a safety block for filtered prompts (15). No idempotency key or safe-retry guidance (0). Only a model and an input are required (7), official SDKs for Python, JavaScript and more (8). 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. Schema & documentation, 56 out of 100, up to 7.2 more on the total Why it scored 56: No OpenAPI or similar contract found for the Interactions endpoint (0). llms.txt with Markdown docs (10). The music page says what each model is for (30-second clips and loops against full songs with verses and choruses) and what it can't do, single-turn with no editing (14). Inputs are a model ID plus text and up to 10 images. Length, structure and instrumental mode are all set in prompt text rather than typed fields (7). Python, JavaScript and REST examples, the safety block described, and Google's shared error model, with no music-specific error list (10). Versioned model IDs and a dated changelog with the Lyria 3.5 GA entry (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. ## 5. Security & auth, 60 out of 100, up to 7 more on the total Why it scored 60: API keys live in a Cloud project, can be restricted to the Gemini API and to IP addresses or apps, the docs give a rotation procedure, and unrestricted standard keys are rejected from 2026-05-28. There are no permission scopes within the API, and the key docs show only the header, no query-string option (25). API restriction keeps a key to the Gemini API, with no read-only mode (5). Returns generated audio and lyrics, no third-party content (10). Usage is visible per project in AI Studio and the Cloud console, and the key docs tell you to audit usage logs (10). security.txt valid on google.com and a public vulnerability reward programme, but we didn't confirm certifications that cover the Gemini API as distinct from Vertex AI (10). 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. ## 6. Transparency & trust, 78 out of 100, up to 1.9 more on the total Made of editorial 55, provenance 100. Why it scored 78: Closed model under the Gemini API terms (modified 2026-04-28). Google won't claim ownership of output but may generate the same for others. SDKs are Apache-2.0 (15). The Lyria pricing rows say paid-tier data isn't used to improve Google products, and since Lyria has no free tier that covers every Lyria call. Abuse-monitoring logs are kept for a limited period, with a number given only for grounding (30 days) (20). The Gemini API deprecations page lists earliest shutdown dates for retiring models, though the Lyria 3 previews were relabelled legacy with no date yet, and price changes take effect 30 days after posting (15). Subprocessors and data locations sit in Google Cloud's terms, which we didn't read for this run (5). 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. ## 7. Maintenance & community, 80 out of 100, up to 1.8 more on the total Why it scored 80: `lyria-3.5` went GA on 2026-09-03, 28 days before this check (30). The Gemini API changelog has many dated entries in the last 90 days, though only one is about Lyria (20). Public changelog and a developer forum, response times not checked (10). Current official SDKs, `google-genai` for Python and `@google/genai` for JavaScript (15). Packages current, CI not checked (5). 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. - Whether a Gemini API SLA or certification scope covers Lyria outside Vertex AI. - The AI Studio status page couldn't be read without a browser. ## Weaknesses - No free tier for Lyria, so billing is needed before the first call - No length parameter, no stems and no multi-turn editing - No rate-limit numbers for Lyria in the docs, and no published contract for the Interactions endpoint - Lyria 3 previews are labelled legacy with no shutdown date yet ## 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. - Draft on `lyria-3-clip-preview` at $0.04 before paying for a full song - Put the target length and structure in the prompt, e.g. "[0:00 - 0:10] Intro" or "a 2-minute song" - Decode `output_audio.data` from base64 straight to a file. Don't let it into the context - Don't name artists or paste song lyrics. The safety filter blocks the request - Back off exponentially on 429 `RESOURCE_EXHAUSTED`. Limits are per project and tier ## What the review panel asked for - A duration field - Rate-limit numbers for Lyria - State blocked-request billing ## 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.