Niches · Voice cloning & custom voices
Train a high-fidelity voice from long recordings
No company's docs show a full fit yet. The listings below do the job with a limit, or not at all.
What counts as done
Docs state the minimum and recommended audio, the steps (upload, speaker separation, verification, training), the training time, and a status the agent can poll or a webhook; training failure returns a reason; the professional voice's price, slot limit and model binding are stated.
How agents ask for it: make a professional clone from my audiobook recordings
, train a voice on 30 minutes of the actor
, highest-quality clone for narration
, fine-tune a voice on our studio recordings
Does the job, with a limit 3
It does the job, but something in the docs or a user's account will get in an agent's way. The limit is named in each row.
| Listing | Why |
|---|---|
| Cartesia Voice Cloning API + MCP C Cartesia AI, Inc. | Pro Voice Clone states 30 minutes minimum, training up to 3 hours, the Startup plan, and an API path (dataset, upload, fine-tune, list voices). No webhook or failure reason found, and slots per plan were not read on the pricing page. source |
| ElevenLabs Voice Cloning and Voice Design API A Eleven Labs Inc. | Docs give 30 minutes minimum and 1 to 3 hours recommended, noise removal, speaker separation, voice verification, and slots (none on Free or Starter, 1 on Creator and Pro, 3 on Scale, 10 on Business). No training-time estimate; the polling field and failure reasons are not in the guide. source |
| Gradium Voice Cloning and Voice Design B Gradium SAS | Pro Voice Clones need 30 minutes minimum (2 hours recommended) and are created only in Gradium Studio with a notification; no API, pollable state or failure reason. source |
Couldn't check 5
The docs were closed to us, or silent. These carry no mark either way.
| Listing | Why |
|---|---|
| Fish Audio Voice Cloning API B Hanabi AI Inc. | The create schema has a pvc_release_state field, but the Professional Voice Clone page with its audio minimum and training time was not read. source |
| Inworld Instant Voice Cloning B Theai, Inc. dba Inworld AI | The guide links Professional Voice Cloning but the page was not read. source |
| Murf Voice Cloning API C Murf Inc. | Murf's terms bar automated access to murf.ai, so its API docs were not read this run, as in the 5 November check; no first-hand account names Murf. |
| Resemble AI Voice Cloning API D Resemble AI, Inc. | Since 30 June 2026 API voices default to rapid clones (listing notes); no professional clone page was found in the docs read. |
| Smallest.ai Instant Voice Clone C Smallest, Inc. | Only the instant clone API was read. |
Doesn't do this job 6
Listings in the category that do something else, with the reason.
| Listing | Why |
|---|---|
| Hume Octave Voice Design and Cloning + MCP C Hume AI, Inc. | Hume's docs index now lists only Audio Decisions and Expression Measurement, and the voice overview and changelog redirect there; the TTS and EVI APIs, cloning and design included, are no longer documented after the 13 November 2026 end of access. source |
| MiniMax Voice Clone and Voice Design C Nanonoble Pte. Ltd. | Only rapid cloning is documented; no professional training tier. source |
| Soniox Voice Cloning BB Soniox Inc. | Instant clones only; no professional training tier. source |
| Speechify API Voice Cloning A Speechify, Inc. | The cloning overview documents zero-shot clones only; no professional training tier. source |
| Ultravox Voice Cloning E Theai, Inc. dba Inworld AI | Instant clones only. source |
| VoiceStudio D VoiceStudio maintainers | No fine-tuning or professional training flow in the MCP tools or README. source |
What stops agents finishing
- one professional slot until the $299 Scale plan (ElevenLabs)
- training failure fixed only by deleting and re-uploading (ElevenLabs help centre)
- pre-Ultra voices cannot generate until upgraded (Resemble)
- voices must be recomputed by hand after a new TTS model ships (Soniox)
Around it
Before: speech-to-text.long-recording-batch, voice-cloning.consented-clone-of-another. After: text-to-speech.long-form-narration.
Where this comes from
Weight 7 of 100 in Voice cloning & custom voices: 2 of 34 accounts (a voice-actor professional clone for a launch, Cartesia's Pro-clone credit rate); the rest rests on ElevenLabs, Cartesia and Resemble docs. Kept as its own case because the steps, the verification and the price differ from instant clones
The same as data: /api/v1/use-cases/voice-cloning.professional-clone-training.json, or the MCP tool get_use_case with id voice-cloning.professional-clone-training. Other jobs in this category: /api/v1/niches?category=voice-cloning.
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.