---
title: Fish Audio S2.1 Pro | Runware Docs
url: https://runware.ai/docs/models/fish-audio-s2-1-pro
description: Flagship multilingual text-to-speech with natural language voice control and realtime streaming
---
# Fish Audio S2.1 Pro

![Fish Audio S2.1 Pro](https://assets.runware.ai/covers/fish-audio-s2-1-pro.jpg)

Fish Audio S2.1 Pro is a flagship text-to-speech model built for highly expressive, low-latency speech generation. It supports natural-language bracket cues for emotion and delivery control, multi-speaker dialogue in a single generation, 80+ languages with automatic language detection, and realtime streaming with very fast time to first audio.

- **Status**: live
- **ID**: `fishaudio:s2.1@pro`
- Try in Playground
- **Creator**: Fish Audio
- **Release Date**: June 1, 2026
- **Capabilities**: Text to Audio, Audio to Audio, Checkpoint

## Pricing

$0.000015 per byte · ~16s per request

**4 example generations**

- [multi-speaker-dialogue](https://runware.ai/docs/models/fish-audio-s2-1-pro/examples#example-midnight-cinema-channel-vocal-stinger): `$0.00356` (~4s)
- [text-to-audio](https://runware.ai/docs/models/fish-audio-s2-1-pro/examples#example-reservoir-science-podcast-cold-open): `$0.0168` (~14s)
- [voice-cloning](https://runware.ai/docs/models/fish-audio-s2-1-pro/examples#example-bronze-bell-inspection-course-narration): `$0.0171` (~18s)
- [text-to-audio](https://runware.ai/docs/models/fish-audio-s2-1-pro/examples#example-forensic-entomology-training-narration): `$0.0227` (~20s)

times vary, measured end to end · [How pricing works](https://runware.ai/docs/models-api/pricing)

## Compatibility & Validation

Inside `speech`, `voice` cannot be used with `voices`.

---

`speech.voice` cannot be used with `inputs.referenceVoices`.

---

`speech.voices` cannot be used with `inputs.referenceVoices`.

## Request Parameters

**API Options**

Platform-level options for task execution and delivery.

### [taskType](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-tasktype)

- **Type**: `string`
- **Required**: true
- **Value**: `audioInference`

Identifier for the type of task being performed.

### [taskUUID](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-taskuuid)

- **Type**: `string`
- **Required**: true
- **Format**: `UUID v4`

UUID v4 identifier for tracking tasks and matching async responses. Must be unique per task.

### [outputType](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-outputtype)

- **Type**: `string`
- **Default**: `URL`

Audio output type.

**Allowed values**: `URL` `base64Data` `dataURI`

### [outputFormat](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-outputformat)

- **Type**: `string`
- **Default**: `MP3`

Specifies the file format of the generated output. The available values depend on the task type and the specific model's capabilities.

- \`MP3\`: Compressed audio, smaller file size.
- \`WAV\`: Uncompressed, high-quality audio.
- \`FLAC\`: Lossless compression.
- \`OGG\`: Open-source compressed audio format (Vorbis codec).

**Allowed values**: `MP3` `WAV` `FLAC` `OGG`

### [audioSettings](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-audiosettings)

- **Path**: `audioSettings.bitrate`
- **Type**: `object (3 properties)`

Audio encoding settings for controlling the bitrate, number of channels, and sample rate of the generated audio. Only applicable for lossy output formats (`MP3` and `OGG`). When using lossless formats (`WAV` or `FLAC`), this parameter must not be provided.

The available sample rates and valid bitrate ranges depend on the output format. For `OGG`, bitrate limits also vary by the number of channels.

**MP3 bitrate limits**:

Bitrate limits for MP3 are the same regardless of mono or stereo.

| Sample Rate | Min Bitrate | Max Bitrate |
| --- | --- | --- |
| 8,000 Hz | 8 kbps | 64 kbps |
| 11,025 Hz | 8 kbps | 64 kbps |
| 12,000 Hz | 8 kbps | 64 kbps |
| 16,000 Hz | 8 kbps | 160 kbps |
| 22,050 Hz | 8 kbps | 160 kbps |
| 24,000 Hz | 8 kbps | 160 kbps |
| 32,000 Hz | 32 kbps | 320 kbps |
| 44,100 Hz | 32 kbps | 320 kbps |
| 48,000 Hz | 32 kbps | 320 kbps |

**OGG bitrate limits — Mono (1 channel)**:

| Sample Rate | Min Bitrate | Max Bitrate |
| --- | --- | --- |
| 8,000 Hz | 8 kbps | 40 kbps |
| 12,000 Hz | 16 kbps | 48 kbps |
| 16,000 Hz | 16 kbps | 96 kbps |
| 24,000 Hz | 16 kbps | 80 kbps |
| 48,000 Hz | 32 kbps | 224 kbps |

**OGG bitrate limits — Stereo (2 channels)**:

| Sample Rate | Min Bitrate | Max Bitrate |
| --- | --- | --- |
| 8,000 Hz | 16 kbps | 80 kbps |
| 12,000 Hz | 16 kbps | 96 kbps |
| 16,000 Hz | 24 kbps | 192 kbps |
| 24,000 Hz | 32 kbps | 160 kbps |
| 48,000 Hz | 48 kbps | 256 kbps |

> [!NOTE]
> **Lossless formats**: When `outputFormat` is set to `WAV` or `FLAC`, the `audioSettings` parameter is not available since these formats produce uncompressed or lossless audio with no configurable encoding settings.

#### [bitrate](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-audiosettings-bitrate)

- **Path**: `audioSettings.bitrate`
- **Type**: `integer`
- **Min**: `8`

Audio bitrate in kbps.

#### [channels](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-audiosettings-channels)

- **Path**: `audioSettings.channels`
- **Type**: `integer`
- **Default**: `2`

Number of audio channels. 1 for mono, 2 for stereo.

**Allowed values**: `1` `2`

#### [sampleRate](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-audiosettings-samplerate)

- **Path**: `audioSettings.sampleRate`
- **Type**: `integer`

Audio sample rate in Hz.

### [webhookURL](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-webhookurl)

- **Type**: `string`
- **Format**: `uri`

Webhook URL that receives JSON responses via HTTP POST when generation tasks complete. For batch requests with multiple results, each completed item triggers a separate webhook call as it becomes available.

**Learn more** (1 resource):

- [Webhooks](https://runware.ai/docs/models-api/webhooks) (platform)

### [deliveryMethod](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-deliverymethod)

- **Type**: `string`
- **Default**: `sync`

Determines how the API delivers task results.

**Allowed values**:

- `sync` Returns results directly in the API response. Over HTTP, a task still running after 120 seconds continues as an async task and comes back with status 'processing'. Poll for its results using getResponse.
- `async` Returns an immediate acknowledgment with the task UUID. Poll for results using getResponse.

**Learn more** (1 resource):

- [Task Polling](https://runware.ai/docs/models-api/task-polling) (platform)

### [uploadEndpoint](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-uploadendpoint)

- **Type**: `string`
- **Format**: `uri`

Specifies a URL where the generated content will be automatically uploaded using the HTTP PUT method. The raw binary data of the media file is sent directly as the request body. For secure uploads to cloud storage, use presigned URLs that include temporary authentication credentials.

**Common use cases:**

- **Cloud storage**: Upload directly to S3 buckets, Google Cloud Storage, or Azure Blob Storage using presigned URLs.
- **CDN integration**: Upload to content delivery networks for immediate distribution.

```text
// S3 presigned URL for secure upload
https://your-bucket.s3.amazonaws.com/generated/content.mp4?X-Amz-Signature=abc123&X-Amz-Expires=3600

// Google Cloud Storage presigned URL
https://storage.googleapis.com/your-bucket/content.jpg?X-Goog-Signature=xyz789

// Custom storage endpoint
https://storage.example.com/uploads/generated-image.jpg
```

The content data will be sent as the request body to the specified URL when generation is complete.

### [ttl](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-ttl)

- **Type**: `integer`
- **Min**: `60`

Time-to-live (TTL) in seconds for generated content. Only applies when `outputType` is `URL`.

### [includeCost](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-includecost)

- **Type**: `boolean`

Include task cost in the response.

### [numberResults](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-numberresults)

- **Type**: `integer`
- **Min**: `1`
- **Max**: `4`
- **Default**: `1`

Number of results to generate. Each result uses a different seed, producing variations of the same parameters.

**Inputs**

Input resources for the task (images, audio, etc). These must be nested inside the \`inputs\` object.

### [referenceVoices](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-inputs-referencevoices)

- **Path**: `inputs.referenceVoices`
- **Type**: `array of objects (2 properties)`
- **Min items**: `items: 1`

Inline voice reference for zero-shot voice cloning. Single-speaker only.

#### [audio](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-inputs-referencevoices-audio)

- **Path**: `inputs.referenceVoices.audio`
- **Type**: `string`
- **Required**: true

Reference audio clip for the voice to clone (UUID, URL, Data URI, or Base64). Audio duration must be between 1.0 and 90.0 seconds.

#### [text](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-inputs-referencevoices-text)

- **Path**: `inputs.referenceVoices.text`
- **Type**: `string`
- **Required**: true
- **Min**: `1`
- **Max**: `1000`

Transcript of the reference audio clip. Must match the spoken content of the reference audio.

**Core Parameters**

Primary parameters that define the task output.

### [model](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-model)

- **Type**: `string`
- **Required**: true
- **Value**: `fishaudio:s2.1@pro`

Send `auto` instead of an AIR identifier to have the model router choose the model for this task. The rest of the request is unchanged, every value you send is kept, and only the parameters you leave out are filled in. See [Model Router](https://runware.ai/docs/models-api/model-router).

**Learn more** (3 resources):

- [Text To Image: Model Selection](https://runware.ai/docs/learn/text-to-image#model-selection) (learn)
- [Image Inpainting: Model Specialized Inpainting Models](https://runware.ai/docs/learn/image-inpainting#model-specialized-inpainting-models) (learn)
- [Image Outpainting: Other Critical Parameters](https://runware.ai/docs/learn/image-outpainting#other-critical-parameters) (learn)

### [speech](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-speech)

- **Path**: `speech.text`
- **Type**: `object (5 properties)`
- **Required**: true

Settings for speech generation.

#### [text](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-speech-text)

- **Path**: `speech.text`
- **Type**: `string`
- **Required**: true
- **Min**: `1`
- **Max**: `3000`

Text to convert to speech. For multi-speaker dialogue, embed inline speaker tags such as `<|speaker:0|>` and `<|speaker:1|>`. Bracket-style paralinguistic cues such as `[laughs]`, `[sigh]`, `[short pause]`, and `[whispering]` can also be embedded inline.

#### [voice](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-speech-voice)

- **Path**: `speech.voice`
- **Type**: `string`

Voice model ID from the Fish Audio voice library or a custom voice. Browse all available voices at https://api.fish.audio/model?page\_size=100&page\_number=1&sort\_by=score.

**Allowed values**:

- `98655a12fa944e26b274c535e5e03842` Egirl, a cute e-girl to chat with you.
- `b347db033a6549378b48d00acb0d06cd` Selene, a meditative female voice.
- `bf322df2096a46f18c579d0baa36f41d` Adrian, a steady and reliable narrator.
- `933563129e564b19a115bedd57b7406a` Sarah, an engaged speaker.
- `536d3a5e000945adb7038665781a4aca` Ethan, a curious explainer.
- `e3cd384158934cc9a01029cd7d278634` Laura, a confident female narrator.
- `79d0bd3e4e5444b18f7b6d89b5927bf1` Jordan, a motivational speaker.
- `9a9cf47702da476aa4629e2506d4a857` Hannah, a conversation specialist.

#### [voices](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-speech-voices)

- **Path**: `speech.voices`
- **Type**: `array of strings`
- **Min items**: `items: 2`

Voice model IDs for multi-speaker dialogue. Speaker tags in `speech.text` such as `<|speaker:0|>` and `<|speaker:1|>` map to the matching array index.

#### [volume](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-speech-volume)

- **Path**: `speech.volume`
- **Type**: `float`
- **Default**: `0`

Volume adjustment in decibels. `0` = no change.

#### [speed](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-speech-speed)

- **Path**: `speech.speed`
- **Type**: `float`
- **Min**: `0.5`
- **Max**: `2`
- **Step**: `0.01`
- **Default**: `1`

Playback speed of the generated speech.

**Settings**

Technical parameters to fine-tune the inference process. These must be nested inside the \`settings\` object.

### [temperature](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-temperature)

- **Path**: `settings.temperature`
- **Type**: `float`
- **Min**: `0`
- **Max**: `1`
- **Step**: `0.01`
- **Default**: `0.7`

Expressiveness of the generated speech.

### [topP](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-topp)

- **Path**: `settings.topP`
- **Type**: `float`
- **Min**: `0.0001`
- **Max**: `1`
- **Step**: `0.01`
- **Default**: `0.7`

Nucleus sampling parameter that controls diversity by limiting the probability mass. Lower values make outputs more focused, higher values increase diversity.

### [chunkLength](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-chunklength)

- **Path**: `settings.chunkLength`
- **Type**: `integer`
- **Min**: `100`
- **Max**: `300`
- **Default**: `300`

Text segment size for processing.

### [conditionOnPreviousChunks](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-conditiononpreviouschunks)

- **Path**: `settings.conditionOnPreviousChunks`
- **Type**: `boolean`
- **Default**: `true`

Previous audio context for voice consistency across chunks.

### [earlyStopThreshold](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-earlystopthreshold)

- **Path**: `settings.earlyStopThreshold`
- **Type**: `float`
- **Min**: `0`
- **Max**: `1`
- **Default**: `1`

Early stopping threshold for batch processing.

### [latency](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-latency)

- **Path**: `settings.latency`
- **Type**: `string`
- **Default**: `normal`

Latency / quality trade-off.

**Allowed values**:

- `low` Fastest response, lower quality.
- `normal` Balanced default.
- `balanced` Highest quality, slower.

### [maxTokens](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-maxtokens)

- **Path**: `settings.maxTokens`
- **Type**: `integer`
- **Min**: `1`
- **Max**: `4294967295`
- **Default**: `1024`

Maximum audio tokens to generate per text chunk.

### [minChunkLength](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-minchunklength)

- **Path**: `settings.minChunkLength`
- **Type**: `integer`
- **Min**: `0`
- **Max**: `100`
- **Default**: `50`

Minimum characters before splitting into a new chunk.

### [normalize](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-normalize)

- **Path**: `settings.normalize`
- **Type**: `boolean`
- **Default**: `true`

Text normalization for English and Chinese. Improves stability for numbers.

### [normalizeLoudness](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-normalizeloudness)

- **Path**: `settings.normalizeLoudness`
- **Type**: `boolean`
- **Default**: `true`

Normalize output loudness for consistent perceived volume.

### [repetitionPenalty](https://runware.ai/docs/models/fish-audio-s2-1-pro#request-settings-repetitionpenalty)

- **Path**: `settings.repetitionPenalty`
- **Type**: `float`
- **Min**: `0`
- **Max**: `2`
- **Step**: `0.01`
- **Default**: `1.2`

Penalty for repeating audio patterns.

## Response Parameters

### [taskType](https://runware.ai/docs/models/fish-audio-s2-1-pro#response-tasktype)

- **Type**: `string`
- **Required**: true
- **Value**: `audioInference`

Identifier for the type of task this response belongs to.

### [taskUUID](https://runware.ai/docs/models/fish-audio-s2-1-pro#response-taskuuid)

- **Type**: `string`
- **Required**: true
- **Format**: `UUID v4`

UUID v4 identifier echoed from the original request, used to match async responses to their tasks.

### [audioUUID](https://runware.ai/docs/models/fish-audio-s2-1-pro#response-audiouuid)

- **Type**: `string`
- **Required**: true
- **Format**: `UUID v4`

UUID of the output audio.

### [audioURL](https://runware.ai/docs/models/fish-audio-s2-1-pro#response-audiourl)

- **Type**: `string`
- **Format**: `uri`

URL of the output audio.

### [audioBase64Data](https://runware.ai/docs/models/fish-audio-s2-1-pro#response-audiobase64data)

- **Type**: `string`

Base64-encoded audio data.

### [audioDataURI](https://runware.ai/docs/models/fish-audio-s2-1-pro#response-audiodatauri)

- **Type**: `string`
- **Format**: `uri`

Data URI of the output audio.

### [cost](https://runware.ai/docs/models/fish-audio-s2-1-pro#response-cost)

- **Type**: `float`

Task cost in USD. Present when `includeCost` is set to `true` in the request.

## Examples

### text-to-audio (text-to-audio)

[Listen to audio](https://assets.runware.ai/examples/fish-audio-s2-1-pro/68b596c8-5600-4e48-81e5-6af933e92084.mp3)

> **Prompt**: In forensic entomology, the insects at a scene are not a clock. They are evidence from which a timeline can be estimated. [short pause] Blow flies may arrive within minutes when remains are accessible, but their development depends on temperature, species, and the conditions surrounding the body. Begin by documenting before collecting. Photograph each insect cluster in place, record its location, and measure the temperature at the body surface, beneath the remains, and within the largest maggot mass. That final reading matters: feeding larvae generate heat, so the center of a mass may be considerably warmer than the surrounding air. Collect specimens from several locations and preserve them in two groups. One group should be preserved promptly for age assessment. The other should remain alive for rearing to adulthood, when species identification is often more reliable. Label every container with the date, time, precise recovery location, and collector's initials. [measured] Weather-station data alone are not sufficient. Investigators must consider shade, indoor climate control, burial depth, wrapping, and any barrier that could delay insect access. These variables can alter colonization and development. [short pause] The key principle is simple: insect evidence does not provide an exact time of death. It establishes a scientifically supported minimum period of colonization. Careful documentation, representative sampling, and an unbroken chain of custody make that estimate defensible.

```json
{
  "taskType": "audioInference",
  "taskUUID": "e17bd5d0-5395-44b3-8aac-f31f63891dc2",
  "model": "fishaudio:s2.1@pro",
  "speech": {
    "text": "In forensic entomology, the insects at a scene are not a clock. They are evidence from which a timeline can be estimated. [short pause] Blow flies may arrive within minutes when remains are accessible, but their development depends on temperature, species, and the conditions surrounding the body.\n\nBegin by documenting before collecting. Photograph each insect cluster in place, record its location, and measure the temperature at the body surface, beneath the remains, and within the largest maggot mass. That final reading matters: feeding larvae generate heat, so the center of a mass may be considerably warmer than the surrounding air.\n\nCollect specimens from several locations and preserve them in two groups. One group should be preserved promptly for age assessment. The other should remain alive for rearing to adulthood, when species identification is often more reliable. Label every container with the date, time, precise recovery location, and collector's initials.\n\n[measured] Weather-station data alone are not sufficient. Investigators must consider shade, indoor climate control, burial depth, wrapping, and any barrier that could delay insect access. These variables can alter colonization and development.\n\n[short pause] The key principle is simple: insect evidence does not provide an exact time of death. It establishes a scientifically supported minimum period of colonization. Careful documentation, representative sampling, and an unbroken chain of custody make that estimate defensible.",
    "voice": "e3cd384158934cc9a01029cd7d278634"
  }
}
```

---

### text-to-audio (text-to-audio)

[Listen to audio](https://assets.runware.ai/examples/fish-audio-s2-1-pro/bf727c25-1f1b-447b-a7dc-4e4fafbb71aa.mp3)

> **Prompt**: [quietly, with intrigue] At 6:12 on a Tuesday morning, the first complaint reached the water department: the tap water tasted like wet soil. By breakfast, forty-three households had called. By noon, people were clearing bottled water from supermarket shelves. [short pause] The reservoir was clean. Every safety test came back normal. But something invisible had crossed the city overnight—and nearly everyone could taste it. [warmer, explanatory] The culprit was geosmin, a harmless compound made by microorganisms and detectable by the human palate in astonishingly tiny amounts. Finding out why it appeared would lead investigators into the chemistry of autumn algae, shifting winds, and a phenomenon called reservoir turnover. [measured emphasis] This is The Water Line, a podcast about the hidden systems behind an ordinary glass of water. I’m your host, and today: how a flavor older than plumbing became a twenty-first-century communications crisis—and what one utility learned from telling the public, truthfully, “It’s safe to drink,” when safe was no longer the only question people were asking.

```json
{
  "taskType": "audioInference",
  "taskUUID": "5e2f5957-4b78-476d-bff8-04cc8e75905c",
  "model": "fishaudio:s2.1@pro",
  "speech": {
    "text": "[quietly, with intrigue] At 6:12 on a Tuesday morning, the first complaint reached the water department: the tap water tasted like wet soil. By breakfast, forty-three households had called. By noon, people were clearing bottled water from supermarket shelves. [short pause] The reservoir was clean. Every safety test came back normal. But something invisible had crossed the city overnight—and nearly everyone could taste it.\n\n[warmer, explanatory] The culprit was geosmin, a harmless compound made by microorganisms and detectable by the human palate in astonishingly tiny amounts. Finding out why it appeared would lead investigators into the chemistry of autumn algae, shifting winds, and a phenomenon called reservoir turnover.\n\n[measured emphasis] This is The Water Line, a podcast about the hidden systems behind an ordinary glass of water. I’m your host, and today: how a flavor older than plumbing became a twenty-first-century communications crisis—and what one utility learned from telling the public, truthfully, “It’s safe to drink,” when safe was no longer the only question people were asking.",
    "voice": "e3cd384158934cc9a01029cd7d278634"
  }
}
```

---

### multi-speaker-dialogue (multi-speaker-dialogue)

[Listen to audio](https://assets.runware.ai/examples/fish-audio-s2-1-pro/75ce51db-d1e5-4a6f-9015-d41294537e05.mp3)

> **Prompt**: <|speaker:0|>[whispering] House lights? <|speaker:1|>Down. [short pause] Projector? <|speaker:0|>Rolling. Tonight’s restoration is ready. <|speaker:1|>[warmly] Then let the lost films speak. Afterimage: midnight cinema, frame by frame.

```json
{
  "taskType": "audioInference",
  "taskUUID": "27e78d19-e4ed-4d94-aded-d60d13c85216",
  "model": "fishaudio:s2.1@pro",
  "speech": {
    "text": "<|speaker:0|>[whispering] House lights?\n<|speaker:1|>Down. [short pause] Projector?\n<|speaker:0|>Rolling. Tonight’s restoration is ready.\n<|speaker:1|>[warmly] Then let the lost films speak. Afterimage: midnight cinema, frame by frame.",
    "voices": [
      "933563129e564b19a115bedd57b7406a",
      "536d3a5e000945adb7038665781a4aca"
    ]
  }
}
```

---

### voice-cloning (voice-cloning)

[Listen to audio](https://assets.runware.ai/examples/fish-audio-s2-1-pro/c0ecdb97-39b9-4d65-8ed9-1e74174a236d.mp3)

> **Prompt**: [calmly] Begin the inspection at the crown, where the bell meets its supporting hardware. Use diffuse light and examine each junction for green powder, lifting scale, or narrow lines radiating into the shoulder. [short pause] A smooth brown or dark green patina is usually stable. Loose, pale green material may indicate active corrosion and should be documented before it is disturbed. Move clockwise around the waist, keeping the same viewing distance. Record dents, casting flaws, and previous repairs separately; they do not carry the same significance. At the sound bow, pay particular attention to bright metal exposed by an uneven clapper strike. Repeated contact outside the established wear track can concentrate stress in the rim. [slightly more serious] Do not perform a sounding test until the supports, clapper, and surrounding structure have been cleared by a qualified engineer. If a crack is visible, suspend all movement of the bell and photograph the feature with a scale marker. [short pause] Your goal is not an immediate diagnosis. It is a consistent record that allows the conservation team to compare change over time.

```json
{
  "taskType": "audioInference",
  "taskUUID": "c4c246b3-d41b-4460-87c7-017b1970ad74",
  "model": "fishaudio:s2.1@pro",
  "speech": {
    "text": "[calmly] Begin the inspection at the crown, where the bell meets its supporting hardware. Use diffuse light and examine each junction for green powder, lifting scale, or narrow lines radiating into the shoulder. [short pause] A smooth brown or dark green patina is usually stable. Loose, pale green material may indicate active corrosion and should be documented before it is disturbed.\n\nMove clockwise around the waist, keeping the same viewing distance. Record dents, casting flaws, and previous repairs separately; they do not carry the same significance. At the sound bow, pay particular attention to bright metal exposed by an uneven clapper strike. Repeated contact outside the established wear track can concentrate stress in the rim.\n\n[slightly more serious] Do not perform a sounding test until the supports, clapper, and surrounding structure have been cleared by a qualified engineer. If a crack is visible, suspend all movement of the bell and photograph the feature with a scale marker. [short pause] Your goal is not an immediate diagnosis. It is a consistent record that allows the conservation team to compare change over time."
  },
  "inputs": {
    "referenceVoices": [
      {
        "audio": "https://assets.runware.ai/assets/inputs/ceea9f17-232a-4640-ba40-e7df23ebdf7f.mp3",
        "text": "My work begins with patience. Bronze records every season it has endured, from winter frost to summer heat. Before any decision is made, I study the surface, note the old repairs, and listen carefully. Conservation is not about making an object new; it is about helping its original voice endure."
      }
    ]
  }
}
```