# Fish Audio TTS API (slim)

> Fish Audio's API turns text into speech with the s2.1-pro model in 83 languages, over a REST endpoint, a timestamped stream and a WebSocket that accepts text as it is produced.

- Full: https://www.anchorterminal.com/tools/fish-audio-tts.md (~7,800 tokens) · this version ~1,680 tokens · JSON https://www.anchorterminal.com/tools/fish-audio-tts.json · canonical https://www.anchorterminal.com/tools/fish-audio-tts
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · Updated: 2026-10-09

**C · 60.7/100 · rank #453 of 842 · #8 in Text-to-speech · not agent-ready · confidence medium**


Assessment: The API accepts streamed text over a WebSocket, returns word timestamps, and prices speech at $15 per million UTF-8 bytes with a $0 model for development. The terms allow training on customer content with no opt-out, and no retention period, SLA document or security certification was found in the reviewed documentation.

## Facts

- Kind: Model API · vendor: Fish Audio · category: Text-to-speech · legal entity: Hanabi AI Inc. · provenance 75/100
- Endpoint: `https://api.fish.audio` (HTTP, Streamable HTTP)
- Auth: OAuth or key · pricing: Pay per use · x402: no · licence: Apache-2.0 (Python SDK), MIT (JavaScript SDK)
- Probe metrics: not measured yet (probes haven't run)
- Models: `s2.1-pro` (default), `s2.1-pro-free`, `s2-pro`, `drama-3-preview` (preview since 23 September 2026), `s1` (deprecated, retires 31 December 2026)
- Languages: 83 on `s2.1-pro`, detected automatically from the text. 13 on `s1`
- Time to first audio: Vendor docs give about 300 ms with `latency` set to `balanced` and 100 ms for `s2-pro`. Not measured by us
- Endpoints: `POST /v1/tts`, `POST /v1/tts/stream/with-timestamp` (SSE), WebSocket `/v1/tts/live` and `/v1/tts/live/with-timestamp`, `GET /model` for the voice library, plus `/compat/v1/audio/speech` and `/compat/elevenlabs`
- Streaming input: The WebSocket takes MessagePack `start`, `text`, `flush` and `stop` events, so LLM tokens can be sent as they arrive
- Speech control: No SSML. Free-form `[bracket]` cues on the S2 family, `prosody.speed` from 0.5 to 2.0, volume in dB, phoneme markup for English, Chinese and Japanese, and up to 3 pronunciation dictionaries a request
- Voices: A `reference_id` from the public Voice Library or your own models, inline reference audio over MessagePack, and several speakers in one request with `<|speaker:N|>` tags
- Output: MP3 at 64, 128 or 192 kbps, WAV or PCM at 8 to 44.1 kHz, Opus at 48 kHz, all mono
- Free tier: `s2.1-pro-free` at $0 under fair-use limits, with no latency or DPA guarantees, free through 30 November 2026 per the changelog
- Rate limits: 5 concurrent requests under $100 prepaid, 15 from $100, 50 from $1,000, shared by all keys on the account. A 429 on the native API has no `Retry-After` header
- Data retention: No fixed period published. The privacy policy keeps Content as long as needed to run the systems, and the compatibility page says zero-retention mode cannot be honoured
- Prices: Text to speech, s2.1-pro $15 per 1M characters; Text to speech, s2.1-pro-free free per 1M characters

- Scores: Reliability 75, Performance pending, Schema & documentation 84, Agent ergonomics 67, Security & auth 35, Payments & pricing 30, Task success pending, Maintenance & community 69, Transparency & trust 60 · total over the 7 assessed categories
- Why: Reliability, Hosted reading. · Schema & documentation, OpenAPI 3.1 file at docs.fish.audio/api-reference/openapi.json, and an AsyncAPI schema rendered on the WebSocket page. · Agent ergonomics, API reading. · Security & auth, Model reading of the checklist, with training and retention in place of the least-privilege and injection lines, as for the other TTS listin… · Payments & pricing, No x402, MPP or L402 (0). · Maintenance & community, Read as a model. · Transparency & trust, Closed API with public terms, SDKs under Apache-2.0 and MIT, and S2-Pro weights published under Fish Audio's own research licence, which is…
- Sources: 29, open questions: 9, both in the full twin
- Capabilities: speech.tts, speech.streaming, speech.voices, speech.languages, voice.clone
- JSON: https://www.anchorterminal.com/api/v1/tools/fish-audio-tts.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/fish-audio-tts.svg` or a link to https://www.anchorterminal.com/tools/fish-audio-tts from a page on fish.audio or one of its subdomains, or the README of github.com/fishaudio/fish-audio-python, then `POST https://www.anchorterminal.com/api/v1/verify` `{"slug", "url"}` or `verify_listing` at /mcp; re-checked weekly, no effect on the grade. Snippets in the full twin.

## Before you call it

1. Send the `model` header on every request and check its spelling. A missing or unknown value is served and billed as `s2.1-pro`.
2. Budget by UTF-8 bytes, not characters. Chinese, Japanese and Korean text costs about three bytes a character.
3. Retry 429 and 5xx with exponential backoff. The native API sends no `Retry-After`, and concurrency starts at 5 for the whole account.
4. Use `[bracket]` cues on the S2 family. `(parenthesis)` tags from `s1` are read aloud as text.
5. Pass the model explicitly in the JavaScript SDK, because npm release 0.1.0 defaults to `s1`.

## Connect

```bash
pip install fish-audio-sdk
```

```bash
curl --request POST https://api.fish.audio/v1/tts \
  --header "Authorization: Bearer $FISH_API_KEY" \
  --header "Content-Type: application/json" \
  --header "model: s2.1-pro-free" \
  --data '{ "text": "Hello from Fish Audio!", "format": "mp3" }' \
  --output hello.mp3
```

```bash
claude mcp add --transport http fish-audio https://api.fish.audio/mcp
```

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| Amazon Polly | BB | 75.6 | speech.tts, speech.streaming, speech.voices, speech.languages | https://www.anchorterminal.com/tools/amazon-polly.min.md |
| ElevenLabs Text to Speech API + MCP | BB | 75.2 | speech.tts, speech.streaming, speech.voices, speech.languages | https://www.anchorterminal.com/tools/elevenlabs-tts.min.md |
| Deepgram Text-to-Speech (Aura-2, Flux TTS) | BB | 72.7 | speech.tts, speech.streaming, speech.voices, speech.languages | https://www.anchorterminal.com/tools/deepgram-tts.min.md |
| Azure AI Speech text-to-speech | BB | 71.4 | speech.tts, speech.streaming, speech.voices, speech.languages | https://www.anchorterminal.com/tools/azure-text-to-speech.min.md |
| Murf TTS API + MCP | BB | 70.6 | speech.tts, speech.streaming, speech.voices, speech.languages | https://www.anchorterminal.com/tools/murf-tts.min.md |


## Panel reviews (0, desk reviews from public material, no calls made)