MiniMax Voice Clone and Voice Design

# Fix list: MiniMax Voice Clone and Voice Design

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/minimax-voice-cloning, the October 2026 research run, assessed 8 October 2026. Grade D, 53.1 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 MiniMax Voice Clone and Voice Design: 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. Security & auth, 39 out of 100, up to 10.7 more on the total

Why it scored 39: Graded for voice cloning, with consent and misuse controls in place of the read-only line and training and retention of voice data in place of the prompt-injection line, as on the other listings in this category. Plain API keys from the console with no scopes found, and the docs say publicly leaked keys may be disabled (20 of 30). Cloning requires individual or enterprise verification of the account, input passes a content safety check, and two optional controls exist, a transcript match (`text_validation`) and a watermark tone (`aigc_watermark`, off by default). No field records the speaker's consent (8 of 20). The terms and privacy policy did not render for our reader, so training use is unknown. The docs state that unused voices are deleted after 7 days and a delete call exists (5 of 15). Responses carry `usage_characters` for reconciliation and voices can be listed, with no audit log found (6 of 15). No security.txt, disclosure policy, bug bounty or certification found, and an issue opened on the MCP repository on 25 August 2026 asks for a private disclosure contact (0 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.

## 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-unit prices are public, $1.50 a cloned voice, $3 a designed voice and $60 to $100 per million characters of speech (20). No free allowance for cloning or design is published. The rate limits page mentions a free tier without defining it (0). A person signs up in a browser, funds a balance and completes verification before the first clone (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, 56 out of 100, up to 7.2 more on the total

Why it scored 56: The clone call returns a preview URL and billing counts, but voice design returns preview audio as hex inside the JSON body, and listing with `voice_type` set to `all` returns the 300+ system voices (15 of 25). `POST /v1/get_voice` filters by voice type and has no pagination (10 of 20). Errors come as numeric codes in `base_resp` with a short remedy for each, several of them specific to cloning, inside a response documented only as 200 (13 of 20). No idempotency key. A repeated clone with the same `voice_id` fails with 2039 instead of creating a second voice, the fee applies only on first synthesis, and the MCP tools carry no read-only or destructive annotations (10 of 20). Two required fields per call. No official SDK for speech, with MCP servers in Python and JavaScript and an official CLI that does not cover cloning (8 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. Reliability, 70 out of 100, up to 6 more on the total

Why it scored 70: Atlassian Statuspage at status.minimax.io with a Text-to-Speech component and no separate component for cloning (20). The page showed incidents from 28 September to 8 October 2026, one of them on Text-to-Speech, 22 minutes of elevated errors on 28 September. The history page did not render and the feed held three days, so the rest of the 90 days was unread, minor only on what we saw (20). Rate limits are published, 60 requests a minute for voice cloning and 20 for voice design (15). Error 1002 and 2045 say to retry later and avoid sudden increases, with no Retry-After, backoff figures or idempotency key found (5 of 15). No SLA found in the docs, and the terms were unread (0). No beta or preview label on either endpoint (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.

## 5. Maintenance & community, 38 out of 100, up to 5.4 more on the total

Why it scored 38: The newest dated change that touches this product is `minimax-mcp` 0.0.19 and `minimax-mcp-js` 0.0.18 on 21 August 2026, 48 days ago. The newest speech model, Speech 2.8, is dated 23 January 2026 (20 of 30). Two package releases for one change in the last 90 days, and no dated API entries (0). The API release notes stop at 28 October 2025. Support is by email and Discord, and the MCP repository lists 21 open issues, whose replies we could not read (6 of 15). No official SDK for speech. The MCP servers were updated in August 2026 and include `voice_clone` and `voice_design` (8 of 15). The Python server has tests and moved to version 2 of the MCP SDK in August, with no CI workflow in the repository (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.

## 6. Transparency & trust, 53 out of 100, up to 4.1 more on the total

Made of editorial 25, provenance 80.

Why it scored 53: Closed API under platform terms that exist but did not render for our reader, with MIT MCP servers (12 of 30). The privacy policy was unread. The docs say every interface is stateless and stores no user data, while also saying an activated voice is saved permanently, and they give no retention period for uploaded audio (5 of 30). The earlier text-to-voice API was marked legacy on 20 June 2025 with service continuing, and legacy speech models are labelled, with no written deprecation policy (8 of 20). No subprocessor list or data locations found. Mainland China accounts use a separate host (0 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: minimax.io, registered 2021-03-27 (5 years) (11 of 15)
- Terms of service: published, but our reader couldn't read it (7 of 10)
- Privacy policy: published, but our reader couldn't read it (7 of 10)
- security.txt: not found (0 of 10)

## 7. Schema & documentation, 78 out of 100, up to 3.6 more on the total

Why it scored 78: OpenAPI 3.1 files for cloning, upload, voice design and voice management (25). llms.txt and a Markdown copy of every docs page (10). Field descriptions state purpose, limits and billing effects, with little on when to choose cloning over design (13 of 20). Enums for model, language and voice type, a 0 to 1 range on `accuracy` and length caps, while the `voice_id` rules sit in prose and the `file_id` description carries a sentence that belongs to `clone_prompt` (12 of 15). Request and response examples throughout and a table of 24 error codes, but only a 200 response is specified and the example response uses fields the schema does not define (10 of 15). Paths are versioned under `/v1`. The API release notes stop at 28 October 2025 and omit `text_validation` and `aigc_watermark` (8 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.

## 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 terms of service, privacy policy and paid agreement at platform.minimax.io/protocol/ return no document text to a plain reader, so voice ownership, the licence taken over uploads, training use, retention, benchmarking rules and any SLA are unread
- unchecked: the legal entity Nanonoble Pte. Ltd. is carried from the 30 September check on the MiniMax Video API listing and was not confirmed today
- unchecked: status history before 28 September 2026. The history page did not render and the feed held three days
- unchecked: GitHub star counts and maintainer replies on issues. The GitHub API refused us for rate limits
- What individual or enterprise verification involves, and whether it applies to voice design as well as cloning
- Which preview price for voice design is right, $30 per million characters in the API reference or $60 on the pricing page
- What the free tier named on the rate limits page includes
- How long uploaded sample files are kept, and how the statement that interfaces store no user data fits with permanently saved voices

## Weaknesses

- The terms of service and privacy policy render only in a browser, so voice ownership, training use and retention terms were not read
- No field records the speaker's consent. The documented gate is individual or enterprise verification of the account
- The voice design reference prices preview audio at $30 per million characters and the pricing page at $60
- The API release notes stop at 28 October 2025 and omit `text_validation` and `aigc_watermark`
- No security.txt, bug bounty or certification found, and no SLA in the docs

## 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.

- Upload the sample to `POST /v1/files/upload` with `purpose=voice_clone`, then pass the integer `file_id` and a new `voice_id` of 8 to 256 characters to `POST /v1/voice_clone`
- Synthesise once with T2A within 7 days or the voice is deleted. A preview inside the clone call does not count
- Read `base_resp.status_code` on every response. Errors such as 2038 (no cloning permission) and 2039 (duplicate `voice_id`) arrive in the body
- Voice design returns `trial_audio` as hex in the JSON body. Write it to a file and keep it out of the model context
- Send `voice_type` to `POST /v1/get_voice`, because `all` also returns the 300+ system voices

## 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.