> ## Documentation Index
> Fetch the complete documentation index at: https://docs.infery.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Audio tools

> Audio tools available as workflow steps and through /v1/capabilities.

**4 tools.** Each one is deterministic: the same input gives the same output, and
nothing is inferred. Call one directly with
[`POST /v1/capabilities/{id}/run`](/api-reference/capabilities/run), or use it as a step in a
[workflow](/workflows/step-types).

Prices are on the [live catalogue](https://infery.ai/models) — they change, and a stale price here would be
worse than none.

| Tool                                 | Does                                                   | In → out      |
| ------------------------------------ | ------------------------------------------------------ | ------------- |
| [`audio.concat`](#audioconcat)       | Join multiple audio clips end-to-end.                  | audio → audio |
| [`audio.pipeline`](#audiopipeline)   | Chain multiple audio operations in one step.           | audio → audio |
| [`audio.transcode`](#audiotranscode) | Re-encode an audio file to a different format/bitrate. | audio → audio |
| [`audio.trim`](#audiotrim)           | Extract a time range from an audio file.               | audio → audio |

## `audio.concat`

**audio → audio** · Billed per second

Join two or more audio clips end-to-end into a single continuous output file.

### When to use

* **Assemble a podcast** — stitch intro, body segments, and outro together.
* **Merge split recordings** — combine parts recorded in separate sessions.
* **Build playlists** — produce a single deliverable file from ordered tracks.

### Inputs & outputs

|                     |                            |
| ------------------- | -------------------------- |
| **Input modality**  | Audio files (list of URLs) |
| **Output modality** | Audio file (GCS URL)       |
| **Min clips**       | 2                          |
| **Max clips**       | 10                         |
| **Max total size**  | 5 GB                       |

### Parameters

No operation-level parameters. All input URLs are supplied as `audios` in the input object.

| Param    | Required | Description                                         |
| -------- | -------- | --------------------------------------------------- |
| *(none)* | —        | Clips are joined in the order provided in `audios`. |

### Examples

#### Join two podcast segments

```json theme={null}
{
  "type": "audio.concat",
  "input": {
    "audios": [
      "https://example.com/example/part1.mp3",
      "https://example.com/example/part2.mp3"
    ]
  },
  "params": {}
}
```

#### Assemble three tracks in order

```json theme={null}
{
  "type": "audio.concat",
  "input": {
    "audios": [
      "https://example.com/example/intro.mp3",
      "https://example.com/example/main.mp3",
      "https://example.com/example/outro.mp3"
    ]
  },
  "params": {}
}
```

### Related capabilities

* `audio.trim` — trim individual clips before concatenating
* `audio.transcode` — re-encode the concatenated output
* `audio.pipeline` — chain concat with other audio ops in one step
* `video.concat` — equivalent operation for video files

### Examples

**Join two clips**

```json theme={null}
{
  "type": "audio.concat",
  "params": {}
}
```

## `audio.pipeline`

**audio → audio** · Billed per second

Chain up to 10 audio operations in a single execution, eliminating intermediate uploads.

### When to use

* **Trim then transcode** — cut a segment and immediately re-encode it.
* **Complex workflows** — multiple dependent steps without round-trip overhead.
* **Atomic processing** — all operations succeed or the whole pipeline fails cleanly.

### Inputs & outputs

|                     |                      |
| ------------------- | -------------------- |
| **Input modality**  | Audio file (URL)     |
| **Output modality** | Audio file (GCS URL) |
| **Max operations**  | 10                   |
| **Max input size**  | 2 GB                 |

### Parameters

| Param        | Required | Description                                                                                                |
| ------------ | -------- | ---------------------------------------------------------------------------------------------------------- |
| `operations` | yes      | Ordered array of 1–10 operation objects. Each object is one of: `{ trim }`, `{ transcode }`, `{ concat }`. |

Each operation uses the same parameter shape as the corresponding standalone capability.

### Examples

#### Trim first 30 s, then convert to MP3

```json theme={null}
{
  "type": "audio.pipeline",
  "input": { "audio_url": "https://example.com/example/recording.wav" },
  "params": {
    "operations": [
      { "trim": { "start": 0, "end": 30 } },
      { "transcode": { "format": "mp3", "bitrate": "192k" } }
    ]
  }
}
```

#### Transcode to AAC at reduced bitrate

```json theme={null}
{
  "type": "audio.pipeline",
  "input": { "audio_url": "https://example.com/example/master.flac" },
  "params": {
    "operations": [
      { "trim": { "start": 10, "end": 120 } },
      { "transcode": { "format": "aac", "bitrate": "128k" } }
    ]
  }
}
```

### Related capabilities

* `audio.transcode` — standalone format conversion
* `audio.trim` — standalone time-range extraction
* `audio.concat` — standalone clip joining
* `video.pipeline` — equivalent multi-op pipeline for video files

### Examples

**Trim then transcode**

```json theme={null}
{
  "type": "audio.pipeline",
  "params": {
    "operations": [
      {
        "trim": {
          "start": 0,
          "end": 30
        }
      },
      {
        "transcode": {
          "format": "mp3"
        }
      }
    ]
  }
}
```

## `audio.transcode`

**audio → audio** · Billed per second

Re-encode an audio file to a different container format and/or bitrate.

### When to use

* **Convert for compatibility** — WAV or FLAC source needs MP3/AAC for web delivery.
* **Reduce file size** — lower bitrate for streaming or podcast distribution.
* **Normalize format** — standardize uploads to a single codec before pipeline processing.

### Inputs & outputs

|                        |                      |
| ---------------------- | -------------------- |
| **Input modality**     | Audio file (URL)     |
| **Output modality**    | Audio file (GCS URL) |
| **Max input duration** | 3 hours              |
| **Max input size**     | 2 GB                 |

### Parameters

| Param     | Required | Description                                                                                                                                                                                  |
| --------- | -------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `format`  | yes      | Output container — `mp3`, `wav`, `aac`, `flac`                                                                                                                                               |
| `bitrate` | no       | Target bitrate as `<int>[k\|M]`, e.g. `192k`, `320k`. Lossy formats only (`mp3`, `aac`) — sending it with `wav` or `flac` is **rejected**, because a lossless encoder has no bitrate to set. |

### Examples

#### WAV to MP3 at 192k

```json theme={null}
{
  "type": "audio.transcode",
  "input": { "audio_url": "https://example.com/example/recording.wav" },
  "params": { "format": "mp3", "bitrate": "192k" }
}
```

#### FLAC to AAC for mobile delivery

```json theme={null}
{
  "type": "audio.transcode",
  "input": { "audio_url": "https://example.com/example/master.flac" },
  "params": { "format": "aac", "bitrate": "128k" }
}
```

### Related capabilities

* `audio.trim` — cut to a time range before transcoding
* `audio.concat` — join clips first, then transcode the combined output
* `audio.pipeline` — chain trim + transcode in one step
* `video.extract_audio` — pull the audio track out of a video first

### Examples

**WAV → MP3 192k**

```json theme={null}
{
  "type": "audio.transcode",
  "params": {
    "format": "mp3",
    "bitrate": "192k"
  }
}
```

## `audio.trim`

**audio → audio** · Billed per second

Extract a specific time range from an audio file, discarding everything outside it.

### When to use

* **Create a clip** — isolate a highlight or excerpt from a long recording.
* **Remove silence** — cut leading/trailing silence from a podcast or voiceover.
* **Prepare for concatenation** — trim individual segments before joining with `audio.concat`.

### Inputs & outputs

|                        |                      |
| ---------------------- | -------------------- |
| **Input modality**     | Audio file (URL)     |
| **Output modality**    | Audio file (GCS URL) |
| **Max input duration** | 3 hours              |
| **Max input size**     | 2 GB                 |

### Parameters

| Param   | Required | Description                                                    |
| ------- | -------- | -------------------------------------------------------------- |
| `start` | yes      | Start time in seconds (non-negative). `0` = beginning of file. |
| `end`   | yes      | End time in seconds (positive, must be > `start`).             |

### Examples

#### Extract the first 60 seconds

```json theme={null}
{
  "type": "audio.trim",
  "input": { "audio_url": "https://example.com/example/podcast.mp3" },
  "params": { "start": 0, "end": 60 }
}
```

#### Extract a mid-section highlight (90–150 s)

```json theme={null}
{
  "type": "audio.trim",
  "input": { "audio_url": "https://example.com/example/podcast.mp3" },
  "params": { "start": 90, "end": 150 }
}
```

### Related capabilities

* `audio.transcode` — re-encode the trimmed output to a different format
* `audio.concat` — join multiple trimmed segments end-to-end
* `audio.pipeline` — combine trim with other audio ops in a single execution
* `video.trim` — equivalent operation for video files

### Examples

**First 60 seconds**

```json theme={null}
{
  "type": "audio.trim",
  "params": {
    "start": 0,
    "end": 60
  }
}
```