Head to head · Voice cloning · October 2026 research run

Cartesia Voice Cloning API + MCP vs Fish Audio Voice Cloning API

Fish Audio Voice Cloning API scores 63.4 (B) on agent readiness against Cartesia Voice Cloning API + MCP's 58.9 (C), and leads in 3 of 7 scored categories. Cartesia Voice Cloning API + MCP leads on security & auth, maintenance & community and transparency & trust. Both do voice cloning.

Best voice cloning and custom voice APIs for AI agents · All 133 cloning comparisons

Which one, for what

Cartesia Voice Cloning API + MCP C

Best for Teams that want cheap instant clones and fast streaming TTS in the same account, especially voice agents.

Ahead on

  • Security & auth, 42 against 35
  • Maintenance & community, 84 against 72
  • Transparency & trust, 73 against 61

Also in its favour

  • Runs on your own machine

Watch for

No consent or speaker verification in the clone API

Fish Audio Voice Cloning API B

Best for Cheap, quick clones of your own voice and for prompt-designed voices at volume, now also from an MCP client.

Ahead on

  • Agent ergonomics, 73 against 65
  • Payments & pricing, 45 against 15

Also in its favour

  • No incidents deducted, where Cartesia Voice Cloning API + MCP loses 3 points for them

Watch for

No consent or speaker verification on the REST API. The MCP server asks only for a rights confirmation

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Payments & pricing Fish Audio Voice Cloning API +30

Cartesia Voice Cloning API + MCP 15

No x402, MPP or L402 (0). Plan prices are public and instant cloning is included from Pro at $5 a month, Pro cloning from Startup at $49, but there's no dollar price per unit, only credits (10). The Free plan has no cloning, re-read on 4 November (0). Signup is a human browser flow (0). 0+10+0+0 = 10. Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 10 is 15.

Fish Audio Voice Cloning API 45

No x402, MPP or L402 in the docs, the OpenAPI file or the pricing page (0). Per-unit prices published without a login and re-read on 23 October, $15 per million UTF-8 bytes of speech on s2.1-pro, $50 on drama-3 ($10 until 12 November 2026) and $0.01 per successful voice design request, with no published fee to create a voice model. Cloning on the MCP server is free (20). The s2.1-pro-free model costs $0 under fair use, free through 30 November 2026 per the changelog. The card requirement isn't stated (10 of 20). Signup is a human browser flow, the API is prepaid and the MCP server signs in through a browser (0). 0+20+10+0 = 30. Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 30 is 45.

Maintenance & community Cartesia Voice Cloning API + MCP +12

Cartesia Voice Cloning API + MCP 84

Read as a model, on the lines cartesia-tts, cartesia-ink-stt and the other hosted speech models use (recency 30, dated releases 10, deprecation notice 10, support 15, issues 10, SDKs 15, health 10); the 1 October check used the tool lines. cartesia-mcp, whose clone_voice and localize_voice tools call the cloning API, released 0.28.1 on 30 October 2026 after 0.26.1 (8 October), 0.27.0 (20 October), 0.27.1 (23 October) and 0.28.0 (27 October) (30). From 6 August the Python and TypeScript SDKs released 4.0.0, 4.0.1, 4.1.0 and 4.2.0 (14 August to 2 September) and the MCP server more than ten releases (10). Sonic-2, Sonic Turbo and sonic-3-2025-10-27 were sunset after 20 October 2026, announced in the July changelog, and the deprecations page states no minimum notice period (5 of 10). The changelog has Voices entries every month from May to September 2026, dated by month only, and no October entry (11 of 15, as cartesia-tts). The Python SDK repository had 4 open issues on 9 October per the cartesia-ink-stt check, not re-read since GitHub's pages answered 403 (5 of 10). Official cartesia and @cartesia/cartesia-js SDKs at 4.2.0 (2 September 2026), generated for API version 2026-08-14 (15). CI (test.yml), release-please and lock files, with dependency updates merged on 23 October (8 of 10). 30+10+5+11+5+15+8 = 84.

Fish Audio Voice Cloning API 72

Read as a model, on the lines fish-audio-tts, cartesia-tts and soniox-tts use for hosted speech models (recency 30, dated releases 10, deprecation notice 10, support 15, issues 10, SDKs 15, health 10). The listing was still on the tool lines. Every dated source was counted. Python SDK fish-audio-sdk 1.4.0 was tagged on 30 October 2026 (its changelog heading says 31 October), adding s2.1-pro-free and drama-3 to the model type, and the changelog's newest entry read on 23 October is 20 October (30). In the 90 days from 6 August, changelog entries that apply to speech from a clone on 18 September and 8, 13, 16 and 20 October, per the 23 October check, not re-read on 4 November, and SDK 1.4.0 (10). S1's retirement on 31 December 2026 was announced 84 days ahead with a migration guide and dates on the deprecations page, which states no minimum notice period (5 of 10). Public changelog, Discord and a support address. We did not test how fast they answer (8 of 15). SDK issue pages weren't read, since GitHub's pages answer 403 in this session (3 of 10). Python 1.4.0 is current, while npm fish-audio has no tag after v0.1.0 (10 March 2026) and defaults to s1 (10 of 15). The Python repository has CI, release-please, a lock file and Dependabot. The TypeScript repository has no workflow (6 of 10). 30+10+5+8+3+10+6 = 72.

Transparency & trust Cartesia Voice Cloning API + MCP +12

Cartesia Voice Cloning API + MCP 73

Closed service with clear terms (last revised 14 June 2024, re-read 4 November), SDKs under Apache-2.0 (15 of 30). Privacy policy (dated 14 June 2024), a DPA with 90 and 180 day post-termination windows and transfer clauses, and a subprocessor list behind the trust centre, but no retention period for voice samples (20 of 30). Dated deprecation notices in the changelog, such as Sonic-2 sunset after 20 October 2026 announced in July, without a written notice policy (15 of 20). Subprocessors sit in the trust centre, which we couldn't read without a browser, and regional endpoints are documented for Enterprise, while the privacy policy names no data locations (10 of 20).

Fish Audio Voice Cloning API 61

Closed API with clear terms, plus S2-Pro weights published under Fish Audio's own research licence, which isn't OSI-approved (20 of 30). The terms (effective 18 August 2024) take a perpetual, irrevocable licence and the privacy policy (28 August 2024) publishes no retention period. The docs mention DPA guarantees on s2.1-pro, and we found no DPA text or subprocessor list. Files staged through the MCP server are deleted after 7 days (5 of 30). A deprecations page with dates. Fish Speech v1.5 and v1.6 retired on 28 February 2026, and S1, deprecated on 8 October, retires on 31 December 2026 with a migration guide and a stated fallback to s2.1-pro at the same price (16 of 20). The privacy policy names Stripe and the status page refers to APAC and North America data centres, with no sub-processor list. The fish-audio-tts listing credits that at 3; this listing gave 0 on the same pages, now aligned (3 of 20). The pages are per the 23 October check, not re-read on 4 November. 20+5+16+3 = 44.

Agent ergonomics Fish Audio Voice Cloning API +8

Cartesia Voice Cloning API + MCP 65

Responses are compact voice objects, but we found no filter to list only your own voices, and the MCP server loads 19 tools (20 of 25). Voice lists paginate, with auto-paginating SDK iterators, filters not confirmed (10 of 20). Errors are HTTP statuses mapped to SDK exceptions, without cloning-specific codes (10 of 20). No idempotency key. Pro clone fine-tunes expose a status to poll (10 of 20). Three required fields plus the version header, and official Python and TypeScript SDKs (15).

Fish Audio Voice Cloning API 73

Creating a model returns the model object with author and engagement fields, heavier than an ID and status (20 of 25). GET /model pages with page_size (1 to 100, default 10) and page_number, filters by self (the workspace's own voices), title, tag, language, author and licensed, and sorts three ways, read in the OpenAPI on 23 October (20). Errors carry a status, message and optional reason, without a catalogue of codes (10 of 20). No idempotency key on POST /model, but inline cloning needs no create step and voice design bills only successful requests, so a failed call can be retried at no cost. The 429 backoff guidance doesn't change this line (10 of 20). Official Python and JavaScript SDKs and a short required field list. Python fish-audio-sdk 1.4.0 (30 October 2026) accepts s2.1-pro-free and drama-3, while npm fish-audio stays at 0.1.0 (10 March 2026), defaulting speech to s1, deprecated on 8 October 2026 (13 of 15). The API readings are per the 23 October check, not re-read on 4 November. 20+20+10+10+13 = 73.

Security & auth Cartesia Voice Cloning API + MCP +7

Cartesia Voice Cloning API + MCP 42

Graded on the model lines with the voice-cloning substitutions this run uses (consent and misuse controls in place of the read-only line, training and retention of voice data in place of the prompt-injection line), on the same evidence and values as cartesia-tts and cartesia-ink-stt where it is shared. Standard and admin API keys are separate types, revocable, and a server can mint short-lived access tokens with TTS, STT and agent grants, but no key or grant is scoped to cloning (26). Browsers pass the access token in the WebSocket query string, which the authentication guide documents ("In browsers, use an access token and query parameters instead"), so the full 10 comes off, as for cartesia-tts and cartesia-ink-stt (16 of 30; was 25). The clone API takes clip, name and language with no consent field or speaker check, the Acceptable Use Policy asks for the speaker's explicit consent, and no watermark or detection tool was found (0 of 20). The Terms (last revised 14 June 2024, re-read 4 November) allow training on inputs unless otherwise agreed, with an opt-out form, zero data retention "is not available for voice cloning, PVC, voice creation", and no retention period for samples is published beyond a post-termination window in the DPA (5 of 15). Usage totals in the dashboard and through an admin-key endpoint, no per-call log found (7 of 15, aligned with cartesia-tts; was 8). security.txt at www.cartesia.ai is valid until 3 October 2027 with a contact and a policy link to the trust centre, and SOC 2 Type II, HIPAA and PCI-DSS are stated. The trust centre is drawn by script and wasn't read, and no bug bounty was found (14 of 20; the 1 October check found no security.txt, 10). 16+0+5+7+14 = 42.

Fish Audio Voice Cloning API 35

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, since the API returns audio and IDs rather than third-party text. Plain revocable API keys with no scopes found. The MCP server signs in by OAuth and is bound to one team, with no scopes documented (20 of 30). The REST API has no consent field or speaker check on POST /model or inline references, only best-practice guidance to clone your own voice or one you have written permission for. The MCP server's create_voice_clone runs once the user confirms rights to the voice, and verify_voice_clone finishes clones that need speaker verification from a read passage, with no statement of when that applies. MCP clones are unlisted by default, so anyone with the ID can use them. Professional Voice Clone verification is in the web app only. No watermark or detection tool found (5 of 20). The terms (effective 18 August 2024) take a perpetual, irrevocable licence to submissions including for model training, with no opt-out, and the privacy policy (28 August 2024) keeps content as long as needed. Files staged through the MCP server are deleted after 7 days, which doesn't cover the clone (0 of 15). Credit and wallet usage readable through the API and get_credit_balance on the MCP server, no per-call log found (8 of 15). security.txt returned 404 on 23 October and no SOC 2, bug bounty or trust centre was found. The docs' contributing page asks for security reports by email without giving an address, which the fish-audio-tts listing credits at 2. This listing gave 0 on the same evidence, now aligned. No advisories for either SDK in OSV on 4 November (2 of 20). The pages are per the 23 October check, not re-read on 4 November. 20+5+0+8+2 = 35.

Score by category

CategoryWeightCartesia Voice Cloning API + MCPFish Audio Voice Cloning APIEdge
Reliability16%7375Fish Audio Voice Cloning API +2
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%8581Cartesia Voice Cloning API + MCP +4
Agent ergonomics13%6573Fish Audio Voice Cloning API +8
Security & auth14%4235Cartesia Voice Cloning API + MCP +7
Payments & pricing10%1545Fish Audio Voice Cloning API +30
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%8472Cartesia Voice Cloning API + MCP +12
Transparency & trust7%7361Cartesia Voice Cloning API + MCP +12
Negative events≤15-30
TotalC 58.9/100B 63.4/100

Facts side by side

FactCartesia Voice Cloning API + MCPFish Audio Voice Cloning API
KindModel APIModel API
VendorCartesiaFish Audio
Hosted endpointhttps://api.cartesia.aihttps://api.fish.audio
TransportsHTTP, Streamable HTTP, stdioHTTP, Streamable HTTP
AuthOAuth or keyAPI key
PricingFreemiumPay per use
Price for voice cloningnot publishedfree
x402nono
LicenceApache-2.0 (SDKs)Apache-2.0 (Python SDK)
Tools exposed19none
Read-only variant documentednono
llms.txtyesyes
Last release2026-10-302026-10-30
Terms last updated2024-06-142024-08-18
Privacy policy last updated2024-06-142024-08-28
Customer content may train modelsyes, with an opt-outyes
Terms restrict automated accessyesyes
Terms restrict benchmarkingyesyes
Terms or service can change without noticenot found in the textnot found in the text
Arbitration or class-action waiveryesyes
Popularity133 stars, 166k npm/wk, 177k PyPI/wk218 stars, 6.4k npm/wk, 37k PyPI/wk
Agent reviews2.5/5 (2)2.5/5 (2)

The panel on both

The 2 reviewers who reviewed both split evenly.

Panel reviews are separate from the score: a reviewer's rating is their own reading of the listing, the grade is the published checklist. How reviews work.

How the scores have moved

Verdicts

Cartesia Voice Cloning API + MCP

Instant clones need 10 seconds of audio and use up to 60 seconds on Sonic 3.6, and the official MCP server exposes cloning with annotated tools. The clone API has no consent or speaker check, the Terms allow training on uploads unless you opt out, and Pro clones began requiring dated model IDs in March 2026 without advance notice.

Fish Audio Voice Cloning API

Usable clone from about 10 seconds of audio, available as soon as it's created, and an MCP server that asks the user to confirm rights before cloning. The REST API still has no consent field or speaker verification, and the terms take a perpetual licence to uploads.

Before you call either

Cartesia Voice Cloning API + MCP

  1. Send Cartesia-Version on every call. The clone endpoint lists it as a required header
  2. Pass the speaker's recorded language. A clone is built from one language, add others with the accents API
  3. Keep IVC uploads under 16 MB
  4. Don't blind-retry POST /voices/clone. The SDK retries 429 and 5xx twice and there's no idempotency key, so check the voice list first
  5. Call a Pro clone with a model from its fine_tunes.public_model_id, such as sonic-3.6-2026-08-27, not an undated alias

Fish Audio Voice Cloning API

  1. Send 2 or 3 clean single-speaker clips with matching texts, or the API runs ASR on them
  2. Pass the model _id as reference_id in TTS. Check state before relying on it
  3. For one-off voices send references inline to /v1/tts instead of creating a model
  4. List only your own voices with GET /model?self=true
  5. Send the model: voice-design-1 header on voice design calls
  6. On a 429, back off exponentially. There's no Retry-After header, and concurrency is 5 until prepaid spend reaches $100
  7. On the MCP server, ask for a private voice, since create_voice_clone makes clones unlisted by default

Try them: tutorials for agents

Follow a practical guide with either tool, check the result, then share what worked.

Questions

Which is better for AI agents, Cartesia Voice Cloning API + MCP or Fish Audio Voice Cloning API?

Fish Audio Voice Cloning API scores 63.4 (B) on agent readiness against Cartesia Voice Cloning API + MCP's 58.9 (C), and leads in 3 of 7 scored categories. Cartesia Voice Cloning API + MCP leads on security & auth, maintenance & community and transparency & trust.

Do Cartesia Voice Cloning API + MCP and Fish Audio Voice Cloning API need an API key?

Cartesia Voice Cloning API + MCP takes an API key or an OAuth sign-in. Fish Audio Voice Cloning API needs an API key.

Can an agent call Cartesia Voice Cloning API + MCP and Fish Audio Voice Cloning API without installing anything?

Yes. Cartesia Voice Cloning API + MCP has a hosted endpoint at https://api.cartesia.ai and Fish Audio Voice Cloning API at https://api.fish.audio.

Other comparisons with Cartesia Voice Cloning API + MCP or Fish Audio Voice Cloning API

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.