# Fix list: Resemble AI Text-to-Speech API From Anchor Terminal's listing at https://www.anchorterminal.com/tools/resemble-ai-tts, the October 2026 research run, assessed 1 October 2026. Grade D, 50.6 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 Resemble AI Text-to-Speech 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, 15 out of 100, up to 10.6 more on the total Why it scored 15: No x402, MPP or L402 (0). Per-second prices published in a public plans API without a login, $0.00067 on Flex and $0.0005 on Team and Business, but the pricing page lists only detection products (15 of 20). No free synthesis allowance. Flex has no fee, then usage is paid from loaded credits (0). A person signs up in a browser (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: Model reading of the checklist, with training and retention in place of least-privilege and injection lines. One Bearer key with no scopes found (20 of 30). The privacy policy (14 August 2026) says customer voice data isn't used to train general-purpose models, worded for voice data rather than text (15 of 20). Voice data is kept while the account is active and for 30 days after it closes, clips are stored in projects, and we found no zero-retention option for synthesis (5 of 15). No per-call log found (0 of 15). A security programme described as aligned with SOC 2 criteria and ISO/IEC 27001 rather than certified, and a security@ address for reports. No security.txt (8 of 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. ## 3. Agent ergonomics, 46 out of 100, up to 8.8 more on the total Why it scored 46: API reading of the checklist. WAV or MP3 from 8 to 48 kHz at 16, 24 or 32-bit, but synchronous synthesis returns the whole clip as base64 inside JSON (18 of 25). 2,000 characters per synchronous or HTTP stream request, 3,000 per WebSocket message. Longer scripts go through projects and clips (10 of 20). Errors are a boolean and a message, with advice to log the request ID (5 of 20). No retry, idempotency or billing-on-failure guidance found (0 of 20). Two required fields, a voice UUID and the text, and SDKs in Node and Python (13 of 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. ## 4. Maintenance & community, 31 out of 100, up to 6 more on the total Why it scored 31: Read as a model. The newest changelog entry is 2026-06-30, 93 days ago (10). No dated entries in the last 90 days (0 of 10). Older models were deprecated on 2026-06-29 and the docs now say voices on them can't generate until upgraded, with no effective date published (0 of 10). Public changelog and a support address, SDK issue trackers not checked (8 of 25). Python SDK 1.9.0 on 2026-04-06 and Node SDK 3.5.3 (10 of 15). The Python SDK still declares support from Python 3.6 (3 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. ## 5. Reliability, 75 out of 100, up to 5 more on the total Why it scored 75: Checkly status page with a Resemble Ultra HTTP check under Text-To-Speech and an activity history (20). One TTS event in the last 90 days, a 1-minute check failure on 22 September, and 100 per cent uptime on that check over 90 days. The page monitors HTTP synthesis only, not the WebSocket (30). 40 requests a second per token published, with WebSocket connection limits in the docs (15). No 429 behaviour or retry guidance found (0). No SLA found (0). Resemble Ultra is the current GA model (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. ## 6. Schema & documentation, 75 out of 100, up to 4.1 more on the total Why it scored 75: OpenAPI file in JSON and YAML at docs.resemble.ai (25). llms.txt with Markdown pages (10). The docs separate synchronous, HTTP streaming and WebSocket synthesis and say WebSocket needs Business, but rarely say when not to use one (12 of 20). Request bodies typed in the spec. The model is chosen by the voice, not the request (12 of 15). Examples on each endpoint, but the errors page only describes `success: false` with a `message` and lists no status codes (6 of 15). `/api/v2` paths and a dated product changelog shared with detection (10 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. ## 7. Transparency & trust, 68 out of 100, up to 2.8 more on the total Made of editorial 50, provenance 86. Why it scored 68: Closed service with clear terms. The SDK licence isn't stated on the listing and we didn't check it (15 of 30). The privacy policy gives retention periods and links a sub-processor list, but names Google Cloud, Render and Cerebrium as infrastructure while the changelog calls Resemble Ultra xAI-powered, and no xAI processing is mentioned (15 of 30). Deprecation started on a dated changelog entry, with no end-of-life date or policy (5 of 20). US hosting with named regions and UK processing for Enterprise on request, sub-processor list at trust.resemble.ai (15 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: resemble.ai, registered 2018-11-12 (7 years) (11 of 15) - security.txt: not found (0 of 10) ## Deductions Each comes off the total. A fixed and documented problem counts for less at the next check. - 2026-06-29. Resemble began deprecating every earlier TTS model, including the Chatterbox family, and the model-versions page now says voices on them can no longer generate audio until upgraded to Resemble Ultra. We found no effective date or notice period. Small deduction because the change is documented and upgrade is self-serve. https://www.resemble.ai/changelog and https://docs.resemble.ai/getting-started/model-versions.md ## 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 xAI processes text sent to Resemble Ultra, and whether it appears on the sub-processor list at trust.resemble.ai, which we couldn't read. - When the older models stopped generating, since no end-of-life date is published. - The patch moves `lastRelease` to the 2026-06-30 changelog entry. The old 2026-04-06 date was the Python SDK release. - How 429 and other failures look on the wire, since the errors page lists no codes. ## Weaknesses - Voices on any pre-Ultra model can't generate until upgraded, with no end-of-life date published - WebSocket streaming only on Business at $1,000 a month - Errors are `success: false` and a message, no status codes listed - Pricing lives in a JSON plans endpoint, not on the pricing page - No changelog entry since 2026-06-30 ## 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. - Check the voice's model before synthesis, since voices on pre-Ultra models fail until upgraded. - Keep each synchronous request under 2,000 characters. - Decode `audio_content` from base64 on `/synthesize`, or call `/stream` for raw WAV chunks. - Send `Authorization: Bearer`, since some doc examples leave out the prefix. - Log the request ID with every failure, since the error body has no code. ## What the review panel asked for - List synthesis on the pricing page - Allow streaming below Business - document 429 behaviour - publish an end-of-life date ## 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.