Gladia

https://docs.gladia.io
Admin
Gladia is a state-of-the-art audio transcription and intelligence API that provides real-time and...
Tokens:136,127
Snippets:893
Trust Score:9.5
Update:4 months ago
Show doc for...
Tokens:
Raw
### Initialize WebSocket for Multi-Channel Audio

Source: https://docs.gladia.io/chapters/live-stt/quickstart

This example demonstrates initializing a single WebSocket connection for multi-channel audio processing. It involves sending a POST request to the Gladia API to get a WebSocket URL and configuring the connection with the correct number of channels, sample rate, and bit depth. Remember to replace '<YOUR_GLADIA_API_KEY>' with your actual API key.

```typescript
// Collect audio buffers from each participant
const samiAudio = getSamiAudioBuffer();
const maximeAudio = getMaximeAudioBuffer();
const markAudio = getMarkAudioBuffer();

// Merge into a multi-channel audio
// Channel ordering: [0]=Sami, [1]=Maxime, [2]=Mark
const channelsData = [samiAudio, maximeAudio, markAudio];
const mergedAudio = interleaveAudio(channelsData, 16); // 16-bit depth

// Initialize a single WebSocket session with multi-channel config
const response = await fetch("https://api.gladia.io/v2/live", {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "x-gladia-key": "<YOUR_GLADIA_API_KEY>",
  },
  body: JSON.stringify({
    encoding: "wav/pcm",
    sample_rate: 16000,
    bit_depth: 16,
    channels: 3, // Specify the number of channels
  }),
});

const { url } = await response.json();
const socket = new WebSocket(url);

// Send the merged audio over a single WebSocket

```

--------------------------------

### Install Gladia SDK for JavaScript

Source: https://docs.gladia.io/chapters/pre-recorded-stt/quickstart

Install the Gladia SDK for JavaScript using npm.

```sh
npm install @gladiaio/sdk
```

--------------------------------

### Start Session Message Example

Source: https://docs.gladia.io/api-reference/v2/live/websocket

This message contains information about the start session of the audio chunk. It is emitted once when the session is created and ready.

```json
{
  "session_id": "550e8400-e29b-41d4-a716-446655440000",
  "created_at": "2025-09-19T12:34:00Z",
  "type": "start_session"
}
```

--------------------------------

### Install Gladia SDK (Python)

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Install the Gladia Python SDK using pip.

```bash
pip install gladiaio-sdk
```

--------------------------------

### Gladia Configuration Example

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Example JSON configuration for a Gladia realtime session, demonstrating model, encoding, language, and other settings.

```json
{
  "model": "solaria-1",
  "encoding": "wav/pcm",
  "bit_depth": 16,
  "sample_rate": 16000,
  "channels": 1,
  "language_config": { "languages": ["en"], "code_switching": false },
  "messages_config": {
    "receive_partial_transcripts": true,
    "receive_final_transcripts": true
  },
  "endpointing": 0.8,
  "maximum_duration_without_endpointing": 30,
  "realtime_processing": {
    "custom_vocabulary": false,
    "custom_spelling": false
  }
}
```

--------------------------------

### Install Gladia SDK (TypeScript)

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Install the Gladia TypeScript SDK using npm.

```bash
npm i @gladiaio/sdk
```

--------------------------------

### Install AssemblyAI SDK (TypeScript)

Source: https://docs.gladia.io/chapters/migrations/from-assembly

Install the AssemblyAI TypeScript SDK using npm.

```bash
npm i assemblyai
```

--------------------------------

### Install AssemblyAI SDK (Python)

Source: https://docs.gladia.io/chapters/migrations/from-assembly

Install the AssemblyAI Python SDK using pip.

```bash
pip install assemblyai
```

--------------------------------

### Speech Start Message Example

Source: https://docs.gladia.io/api-reference/v2/live/websocket

Indicates the start of detected speech activity on a given channel. Includes session ID, timestamp, and channel information.

```json
{
  "session_id": "550e8400-e29b-41d4-a716-446655440000",
  "created_at": "2025-09-19T12:34:06Z",
  "type": "speech_start",
  "data": {
    "time": 1.24,
    "channel": 0
  }
}
```

--------------------------------

### Start Session

Source: https://docs.gladia.io/api-reference/v2/live/websocket

Emitted once when the session is created and ready. This message contains the information about the start session of the audio chunk.

```APIDOC
## Start Session

### Description
This message contains the information about the start session of the audio chunk. It is emitted once when the session is created and ready.

### Method
SEND

### Endpoint
N/A (WebSocket Event)

### Parameters
#### Payload
- **session_id** (string) - Required - UUID identifying the session.
- **created_at** (string) - Required - ISO 8601 timestamp in UTC.
- **type** (string) - Required - Lifecycle event emitted once when the session is created and ready.

### Request Example
```json
{
  "session_id": "550e8400-e29b-41d4-a716-446655440000",
  "created_at": "2025-09-19T12:34:00Z",
  "type": "start_session"
}
```

### Response
N/A (This is a message sent by the server)

```

--------------------------------

### Install Gladia n8n Node via npm

Source: https://docs.gladia.io/chapters/integrations/n8n

Install the Gladia community node for n8n using npm. This is required for self-hosted n8n instances.

```bash
npm install @gladiaio/n8n-nodes
```

--------------------------------

### Start recording callback

Source: https://docs.gladia.io/llms.txt

Defines the payload for the callback event `live.start_recording`.

```APIDOC
## POST /v2/live/callback/start-recording

### Description
Payload definition for the callback event `live.start_recording`.

### Method
POST

### Endpoint
/v2/live/callback/start-recording
```

--------------------------------

### Start recording webhook

Source: https://docs.gladia.io/llms.txt

Defines the payload for the webhook event `live.start_recording`.

```APIDOC
## POST /v2/live/webhook/start-recording

### Description
Payload definition for the webhook event `live.start_recording`.

### Method
POST

### Endpoint
/v2/live/webhook/start-recording
```

--------------------------------

### V2 Upload Response Example

Source: https://docs.gladia.io/chapters/pre-recorded-stt/migration-from-v1

This is an example of the JSON response you will receive after successfully uploading an audio file using the V2 upload endpoint. It contains the audio_url and metadata needed for subsequent requests.

```json
{
  "audio_url": "https://api.gladia.io/file/636c70f6-92c1-4026-a8b6-0dfe3ecf826f",
  "audio_metadata": {
    "id": "636c70f6-92c1-4026-a8b6-0dfe3ecf826f",
    "filename": "conversation.wav",
    "extension": "wav",
    "size": 99515383,
    "audio_duration": 4146.468542,
    "number_of_channels": 2
  }
}
```

--------------------------------

### Install Deepgram SDK (TypeScript)

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Install the Deepgram TypeScript SDK using npm.

```bash
npm i @deepgram/sdk
```

--------------------------------

### Start session callback

Source: https://docs.gladia.io/llms.txt

Defines the payload for the callback event `live.start_session`.

```APIDOC
## POST /v2/live/callback/start-session

### Description
Payload definition for the callback event `live.start_session`.

### Method
POST

### Endpoint
/v2/live/callback/start-session
```

--------------------------------

### Install Deepgram SDK (Python)

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Install the Deepgram Python SDK using pip.

```bash
pip install deepgram-sdk
```

--------------------------------

### Example Post-Meeting Analysis Response

Source: https://docs.gladia.io/chapters/audio-intelligence/audio-to-llm

An example of the structured response received after processing audio with the configured LLM prompts. It includes success status, execution times, and detailed results for each prompt.

```json
{
  "success": true,
  "is_empty": false,
  "results": [
    {
      "success": true,
      "is_empty": false,
      "results": {
        "prompt": "Summarize the meeting as bullet points: main topics, decisions, and open questions.",
        "response": "- **Roadmap Q2**: Team aligned on shipping the billing integration first.\n- **Decision**: Weekly sync moved to Tuesday.\n- **Open question**: Whether to support SSO in v1 is still TBD."
      },
      "exec_time": 1.7726809978485107,
      "error": null
    },
    {
      "success": true,
      "is_empty": false,
      "results": {
        "prompt": "Give a concise paragraph summarizing what this meeting was about and the outcome.",
        "response": "The group reviewed Q2 priorities, agreed to prioritize billing, and rescheduled the standing meeting. SSO scope was left for a follow-up once design signs off."
      },
      "exec_time": 1.5122809978485107,
      "error": null
    },
    {
      "success": true,
      "is_empty": false,
      "results": {
        "prompt": "List action items and follow-ups to prepare for the next meeting; include owners if they were mentioned.",
        "response": "- **Alex**: Finalize SSO requirements doc by Friday.\n- **Jamie**: Share billing API cutover checklist with the team.\n- **Everyone**: Review the updated roadmap draft before next sync."
      },
      "exec_time": 1.8932809978258485,
      "error": null
    }
  ],
  "exec_time": 6.267103805541992,
  "error": null
}
```

--------------------------------

### Start Gladia Transcription Session (Python)

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Initiates a live transcription session with Gladia using the Python SDK. Ensure the gladia_config is properly set up.

```python
gladia_session = gladia_client.live_v2().start_session(gladia_config)
```

--------------------------------

### Start session webhook

Source: https://docs.gladia.io/llms.txt

Defines the payload for the webhook event `live.start_session`.

```APIDOC
## POST /v2/live/webhook/start-session

### Description
Payload definition for the webhook event `live.start_session`.

### Method
POST

### Endpoint
/v2/live/webhook/start-session
```

--------------------------------

### Example Custom Metadata Structure

Source: https://docs.gladia.io/chapters/audio-intelligence/custom-metadata

This JSON structure shows an example of how to format custom metadata. It can include various data types for flexible tagging.

```json
{
  "custom_metadata": {
    "internalUserId": 2348739875894375,
    "paymentMethod": { "last4Digits": 4576 },
    "internalUserName": "Spencer"
  }
}
```

--------------------------------

### Start Gladia Transcription Session (TypeScript)

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Initiates a live transcription session with Gladia using the TypeScript SDK. Ensure the gladiaConfig is properly set up.

```typescript
const gladiaSession = gladiaClient.liveV2().startSession(gladiaConfig);
```

--------------------------------

### Initiate Live STT Session (Python)

Source: https://docs.gladia.io/chapters/live-stt/quickstart

Initiates a live STT session with specified configurations for messages and callbacks. This is the first step to start transcribing audio in real-time.

```python
from gladia import LiveClient

# Replace with your actual API key
client = LiveClient("YOUR_GLADIA_API_KEY")

# Configure messages and callbacks as needed
# See API reference for all options
config = {
    "language": "en",
    "messages_config": {
        "receive_transcript": True,
        "receive_partial_transcripts": True
    },
    "callback_config": {
        "urls": ["YOUR_CALLBACK_URL"]
    }
}

# Start the session
live_session = client.create_live_session(**config)

print(f"Session started. Listening...")
```

--------------------------------

### Start Recording Message Schema

Source: https://docs.gladia.io/api-reference/v2/live/websocket

Defines the structure for messages indicating the start of an audio recording. Includes session ID, creation timestamp, and event type.

```json
{
  "session_id": "550e8400-e29b-41d4-a716-446655440000",
  "created_at": "2025-09-19T12:34:05Z",
  "type": "start_recording"
}
```

--------------------------------

### Speech Start callback

Source: https://docs.gladia.io/llms.txt

Defines the payload for the callback event `live.speech_start`.

```APIDOC
## POST /v2/live/callback/speech-start

### Description
Payload definition for the callback event `live.speech_start`.

### Method
POST

### Endpoint
/v2/live/callback/speech-start
```

--------------------------------

### Start Recording Event

Source: https://docs.gladia.io/api-reference/v2/live/websocket

This message indicates the beginning of an audio recording session. It includes a unique session ID, a timestamp, and the event type.

```APIDOC
## Start Recording Event

### Description
This message contains information about the start of an audio chunk recording.

### Event Type
`start_recording`

### Payload Schema
```json
{
  "type": "object",
  "required": [
    "session_id",
    "created_at",
    "type"
  ],
  "properties": {
    "session_id": {
      "type": "string",
      "description": "UUID identifying the session."
    },
    "created_at": {
      "type": "string",
      "description": "ISO 8601 timestamp in UTC."
    },
    "type": {
      "type": "string",
      "const": "start_recording",
      "description": "Lifecycle event emitted when audio capture/streaming begins."
    }
  }
}
```

### Example
```json
{
  "session_id": "550e8400-e29b-41d4-a716-446655440000",
  "created_at": "2025-09-19T12:34:05Z",
  "type": "start_recording"
}
```
```

--------------------------------

### Live Translation Configuration

Source: https://docs.gladia.io/chapters/audio-intelligence/translation

Example of enabling translation for live audio streams. This involves configuring the realtime processing and message reception.

```APIDOC
## Live Translation Configuration

### Description
Enables translation for live audio streams. This configuration is part of the `realtime_processing` object and includes similar translation settings as pre-recorded audio, along with a setting to receive real-time processing events.

### Method
(Not specified in source, likely part of a WebSocket connection or streaming API)

### Endpoint
(Not specified in source)

### Parameters
#### Request Body (within `realtime_processing` and `messages_config`)
- **realtime_processing.translation** (boolean) - Required - Set to `true` to enable translation for the live stream.
- **realtime_processing.translation_config** (object) - Required - Configuration object for translation.
  - **target_languages** (string[]) - Required - Array of language codes for translation output (e.g., ["fr"]).
  - **model** (enum["base", "enhanced"]) - Optional - Specifies the translation model to be used. Defaults to "base".
  - **match_original_utterances** (boolean) - Optional - Keep translated segments aligned with source segmentation. Defaults to `true`.
  - **lipsync** (boolean) - Optional - Controls alignment with visual cues (lip movements). Defaults to `true`.
  - **context_adaptation** (boolean) - Optional - Enable context-aware translation. Defaults to `true`.
  - **context** (string) - Optional - Additional context to improve terminology, proper nouns, or disambiguation. Effective with `context_adaptation: true`.
  - **informal** (boolean) - Optional - Prefer informal register when available. Defaults to `false`.
- **messages_config.receive_realtime_processing_events** (boolean) - Required - Set to `true` to receive real-time processing events, including translations.

### Request Example
```json
{
  "realtime_processing": {
    "translation": true,
    "translation_config": {
      "target_languages": [
        "fr"
      ],
      "model": "base",
      "match_original_utterances": true,
      "lipsync": true,
      "context_adaptation": true,
      "context": "<string>",
      "informal": false
    }
  },
  "messages_config": {
    "receive_realtime_processing_events": true
  }
}
```

### Response
(Response structure not detailed in source)
```

--------------------------------

### Initiate Live Session with GladiaClient in Python

Source: https://docs.gladia.io/chapters/live-stt/quickstart

Initialize the GladiaClient and create a live session configuration object in Python. This prepares the client for real-time audio processing with specified parameters.

```python
# Our Python SDK supports sync/threaded and asyncio versions.
gladi_client = GladiaClient(api_key="<YOUR_GLADIA_API_KEY>")

# sync/threaded version
live_client = gladia_client.live_v2()
# asyncio version
live_client = gladia_client.live_v2_async()

init_request = LiveV2InitRequest(
    model="solaria-1",
    encoding="wav/pcm",
    sample_rate=16000,
    bit_depth=16,
    channels=1,
    language_config=LiveV2LanguageConfig(languages=["fr"], code_switching=False),
    messages_config=LiveV2MessagesConfig(receive_partial_transcripts=True),
)

live_session = live_client.start_session(init_request)
```

--------------------------------

### Start Deepgram Transcription Session (Python)

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Initiates a live transcription session with Deepgram using the Python SDK. Ensure the deepgram_config is properly set up.

```python
with deepgram_client.listen.v1.connect(deepgram_config) as deepgram_connection:
    # Further code to add event handlers
    deepgram_connection.start_listening()
```

--------------------------------

### Start Deepgram Transcription Session (TypeScript)

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Initiates a live transcription session with Deepgram using the TypeScript SDK. Ensure the deepgramConfig is properly set up.

```typescript
const deepgramConnection = deepgramClient.listen.live(deepgramConfig);
```

--------------------------------

### Pre-recorded Translation Configuration

Source: https://docs.gladia.io/chapters/audio-intelligence/translation

Example of enabling translation for pre-recorded audio. This includes specifying target languages, the translation model, and various alignment and context options.

```APIDOC
## Pre-recorded Translation Configuration

### Description
Enables translation for pre-recorded audio content. You can specify multiple target languages, choose between 'base' and 'enhanced' translation models, and fine-tune parameters like utterance matching, lip-sync, and context adaptation.

### Method
POST (assumed, based on request body structure)

### Endpoint
(Not specified in source, typically part of a transcription or processing endpoint)

### Parameters
#### Request Body
- **translation** (boolean) - Required - Set to `true` to enable translation.
- **translation_config** (object) - Required - Configuration object for translation.
  - **target_languages** (string[]) - Required - Array of language codes for translation output (e.g., ["fr"]).
  - **model** (enum["base", "enhanced"]) - Optional - Specifies the translation model to be used. Defaults to "base".
  - **match_original_utterances** (boolean) - Optional - Keep translated segments aligned with source segmentation. Defaults to `true`.
  - **lipsync** (boolean) - Optional - Controls alignment with visual cues (lip movements). Defaults to `true`.
  - **context_adaptation** (boolean) - Optional - Enable context-aware translation. Defaults to `true`.
  - **context** (string) - Optional - Additional context to improve terminology, proper nouns, or disambiguation. Effective with `context_adaptation: true`.
  - **informal** (boolean) - Optional - Prefer informal register when available. Defaults to `false`.

### Request Example
```json
{
  "translation": true,
  "translation_config": {
    "target_languages": [
      "fr"
    ],
    "model": "base",
    "match_original_utterances": true,
    "lipsync": true,
    "context_adaptation": true,
    "context": "<string>",
    "informal": false
  }
}
```

### Response
(Response structure not detailed in source)
```

--------------------------------

### Example Sentence Segmentation Result

Source: https://docs.gladia.io/chapters/pre-recorded-stt/features/sentences

The API response will include a 'sentences' key containing segmented results, including sentence text, start and end times, confidence, language, speaker, and channel.

```json
"sentences": {
  "success": true,
  "is_empty": false,
  "results": [
    {
      "sentence": "Amy, it says you are trained in technology.",
      "start": 0.4681999999999999,
      "end": 2.45525,
      "words": [...],
      "confidence": 0.95,
      "language": "en",
      "speaker": 0,
      "channel": 0
    }
  ]
}
```

--------------------------------

### V2 Full Configuration Sample

Source: https://docs.gladia.io/chapters/live-stt/migration-from-v1

A comprehensive V2 configuration demonstrating the reorganized parameters for language, preprocessing, and realtime processing.

```json
{
  "encoding": "wav/pcm",
  "bit_depth": 8,
  "sample_rate": 48000,
  "endpointing": 0.8,
  "maximum_duration_without_endpointing": 10,
  "language_config": {
    "languages": ["en"]
  }
  "pre_processing": {
    "audio_enhancer": true
  },
  "realtime_processing": {
    "words_accurate_timestamps": true
  }
}
```

--------------------------------

### Audio-to-LLM Response Structure for One Prompt

Source: https://docs.gladia.io/chapters/audio-intelligence/audio-to-llm

Example response shape from an Audio-to-LLM request when a single prompt is used. It includes the prompt, the LLM's response, and execution details.

```json
{
  "success": true,
  "is_empty": false,
  "results": [
    {
      "success": true,
      "is_empty": false,
      "results": {
        "prompt": "Summarize the transcript in three bullet points.",
        "response": "- Intro and context\n- Main discussion\n- Conclusion and next steps"
      },
      "exec_time": 1.4122809978485107,
      "error": null
    }
  ],
  "exec_time": 4.521103805541992,
  "error": null
}
```

--------------------------------

### Filtering Live Transcriptions with Multiple Metadata Fields

Source: https://docs.gladia.io/chapters/audio-intelligence/custom-metadata

This example demonstrates filtering live transcriptions using multiple custom metadata fields, including nested objects.

```bash
https://api.gladia.io/v2/live?custom_metadata={"paymentMethod": {"last4Digits": 4576}, "internalUserName": "Spencer"}
```

--------------------------------

### Initiate Live Job

Source: https://docs.gladia.io/api-reference/v2/live/init

Initiates a new live audio transcription job. This endpoint allows you to start a real-time transcription session by providing audio stream configuration details.

```APIDOC
## POST /v2/live

### Description
Initiates a new live job for audio processing.

### Method
POST

### Endpoint
https://api.gladia.io/v2/live

### Parameters
#### Query Parameters
- **region** (string) - Optional - The region used to process the audio. Supported values: `us-west`, `eu-west`.

#### Request Body
- **encoding** (string) - Optional - The encoding format of the audio stream. Supported formats: PCM (8, 16, 24, 32 bits), A-law (8 bits), μ-law (8 bits). Defaults to `wav/pcm`.
- **bit_depth** (integer) - Optional - The bit depth of the audio stream. Defaults to `16`.
- **sample_rate** (integer) - Optional - The sample rate of the audio stream. Defaults to `16000`.
- **channels** (integer) - Optional - The number of channels of the audio stream. Minimum: `1`, Maximum: `8`. Defaults to `1`.
- **custom_metadata** (object) - Optional - Custom metadata you can attach to this live transcription.
- **model** (string) - Optional - The model used to process the audio. Defaults to `solaria-1`.
- **endpointing** (number) - Optional - The endpointing duration in seconds. Defaults to `0.05`.
- **maximum_duration_without_endpointing** (number) - Optional - The maximum duration in seconds without endpointing. Defaults to `5`.
- **language_config** (object) - Optional - Specify the language configuration.
- **pre_processing** (object) - Optional - Specify the pre-processing configuration.
- **realtime_processing** (object) - Optional - Specify the realtime processing configuration.
- **post_processing** (object) - Optional - Specify the post-processing configuration.
- **messages_config** (object) - Optional - Specify the websocket messages configuration.
- **callback** (boolean) - Optional - If true, messages will be sent to configured url. Defaults to `false`.

### Responses
#### Success Response (201)
- **live_job_id** (string) - The ID of the initiated live job.

#### Error Responses
- **400** - Bad Request: Something is wrong with the request.
- **401** - Unauthorized: You don't have the permissions to initiate a new live job.
- **422** - Unprocessable Entity: The parameters you gave are incorrect.
```

--------------------------------

### Initialize Gladia Client (Python)

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Initialize the Gladia client with your API key in Python.

```python
from gladiaio_sdk import GladiaClient

gladiaclient = GladiaClient(api_key="<YOUR_GLADIA_API_KEY>")
```

--------------------------------

### Initiate Live Session with GladiaClient in JavaScript

Source: https://docs.gladia.io/chapters/live-stt/quickstart

Create a GladiaClient instance and configure a live session with specified audio parameters and language settings. This sets up the client for real-time transcription.

```typescript
const gladiaClient = new GladiaClient({
  apiKey: <YOUR_GLADIA_API_KEY>,
});

const gladiaConfig = {
  model: "solaria-1",
  encoding: 'wav/pcm',
  sample_rate: 16000,
  bit_depth: 16,
  channels: 1,
  language_config: {
    languages: ["fr"],
    code_switching: false,
  },
};

const liveSession = gladiaClient.liveV2().startSession(gladiaConfig);
```

--------------------------------

### V1 Full Configuration Sample

Source: https://docs.gladia.io/chapters/live-stt/migration-from-v1

A comprehensive V1 configuration including language, audio enhancement, and timestamp settings.

```json
{
  "encoding": "wav",
  "bit_depth": 8,
  "sample_rate": 48000,
  "model": "accurate",
  "endpointing": 800,
  "maximum_audio_duration": 10,
  "language_behaviour": "manual",
  "language": "english",
  "audio_enhancer": true,
  "word_timestamps": true
}
```

--------------------------------

### Make Integration

Source: https://docs.gladia.io/llms.txt

How to use Gladia with Make.

```APIDOC
## Make Integration

### Description
How to use Gladia with Make.
```

--------------------------------

### Audio Chunk Acknowledgment Example

Source: https://docs.gladia.io/api-reference/v2/live/websocket

Example of an acknowledgment message for a received audio chunk. It confirms successful reception and buffering by the server.

```json
{
  "session_id": "550e8400-e29b-41d4-a716-446655440000",
  "created_at": "2025-09-19T12:47:00Z",
  "acknowledged": true,
  "type": "audio_chunk_ack",
  "data": {
    "byte_range": [
      0,
      1024
    ],
    "time_range": [
      0.0,
      1.024
    ]
  }
}
```

--------------------------------

### Initialize AssemblyAI Client (Python)

Source: https://docs.gladia.io/chapters/migrations/from-assembly

Initialize the AssemblyAI streaming client with your API key and host.

```python
from assemblyai.streaming.v3 import StreamingClient, StreamingClientOptions

api_key = "<YOUR_ASSEMBLYAI_API_KEY>"

assemblyClient = StreamingClient(
    StreamingClientOptions(
        api_key=api_key,
        api_host="streaming.assemblyai.com",
    )
)
```

--------------------------------

### Post Summarization Message Example

Source: https://docs.gladia.io/api-reference/v2/live/websocket

Example of a message containing post-processing abstractive summary of a conversation. This is used after audio processing is complete.

```json
{
  "session_id": "550e8400-e29b-41d4-a716-446655440000",
  "created_at": "2025-09-19T12:47:00Z",
  "type": "post_summarization",
  "error": null,
  "data": {
    "results": "The team aligned on goals and next steps for the quarter."
  }
}
```

--------------------------------

### V2 Transcription Message Example

Source: https://docs.gladia.io/chapters/live-stt/migration-from-v1

Example of a final transcription message received from the V2 API. It uses 'type': 'transcript' and 'data.is_final': true for identification.

```json
{
  "type": "transcript",
  "session_id": "de0a341d-c69f-4e15-a649-7b3f49e211f0",
  "created_at": "2024-10-10T14:35:28.387Z",
  "data": {
    "id": "00_00000000",
    "is_final": true,
    "utterance": {
      "text": " Hello world",
      "start": 0.188,
      "end": 1.284,
      "language": "en",
      "confidence": 1,
      "channel": 0,
      "words": [
        { "word": " Hello", "start": 0.188, "end": 0.735, "confidence": 1 },
        { "word": " world", "start": 0.736, "end": 1.284, "confidence": 1 }
      ]
    }
  }
}
```

--------------------------------

### V1 Transcription Message Example

Source: https://docs.gladia.io/chapters/live-stt/migration-from-v1

Example of a final transcription message received from the V1 API. It includes fields like 'event', 'type', and 'transcription'.

```json
{
  "event": "transcript",
  "request_id": "G-3abade39",
  "type": "final",
  "transcription": " Hello world",
  "time_begin": 1.4376875,
  "time_end": 2.4696875,
  "confidence": 0.65,
  "language": "en",
  "utterances": [
    {
      "transcription": " Hello world",
      "time_begin": 1.4376875,
      "time_end": 2.4696875,
      "language": "en",
      "confidence": 0.65,
      "stable": true,
      "id": 0
    }
  ],
  "inference_time": 0.843909502029419,
  "duration": 2.5845000000000002
}
```

--------------------------------

### Custom Spelling Example Dictionary

Source: https://docs.gladia.io/chapters/audio-intelligence/custom-spelling

An example of a custom spelling dictionary where keys are the correct spellings and values are the variants to be replaced. This is used for literal matching.

```json
{
  "Gorish": ["ghorish", "gaurish", "gaureish", "geurish", "go rich"]
}
```

--------------------------------

### Configuration Comparison: Endpointing and Max Duration (V1 vs V2)

Source: https://docs.gladia.io/chapters/live-stt/migration-from-v1

Illustrates the change in configuration for endpointing and maximum audio duration between V1 and V2. V2 uses seconds for endpointing and a different key for maximum duration.

```json
{
  "endpointing": 800,
  "maximum_audio_duration": 10
}
```

```json
{
  "endpointing": 0.8,
  "maximum_duration_without_endpointing": 10
} 
```

--------------------------------

### Initialize Gladia Client (TypeScript)

Source: https://docs.gladia.io/chapters/migrations/from-deepgram

Initialize the Gladia client with your API key in TypeScript.

```typescript
import { GladiaClient } from "@gladiaio/sdk";

const gladiaClient = new GladiaClient({
  apiKey: process.env.GLADIA_API_KEY,
});
```

--------------------------------

### Meeting BaaS Integration

Source: https://docs.gladia.io/llms.txt

How to integrate Gladia with Meeting BaaS.

```APIDOC
## Meeting BaaS Integration

### Description
How to integrate Gladia with Meeting BaaS.
```

--------------------------------

### Initiate WebSocket Connection (V1)

Source: https://docs.gladia.io/chapters/live-stt/migration-from-v1

Connect to the V1 WebSocket URL and send configuration. Requires the 'ws' library.

```javascript
import WebSocket from 'ws';

const socket = new WebSocket('wss://api.gladia.io/audio/text/audio-transcription');

socket.addEventListener("open", function() {
  // Send configuration
  socket.send(JSON.stringify({
    'x_gladia_key': 'YOUR_GLADIA_API_KEY',
    // ...config properties
  }))

  // Start sending audio chunks
});
```

--------------------------------

### Get Transcription Job Metadata (OpenAPI)

Source: https://docs.gladia.io/api-reference/v2/transcription/get

This OpenAPI definition outlines the GET request to retrieve transcription job metadata. It specifies the endpoint, parameters, and possible responses, including success (200) and error codes (401, 404).

```yaml
openapi: 3.1.0
info:
  title: Gladia Control API
  description: ''
  version: '1.0'
  contact: {}
servers:
  - url: https://api.gladia.io/
    description: Gladia API production URL
security: []
tags: []
paths:
  /v2/transcription/{id}:
    get:
      tags:
        - Transcription V2
      summary: Get the transcription job's metadata
      operationId: TranscriptionController_getTranscript_v2
      parameters:
        - name: id
          required: true
          in: path
          description: Id of the transcription job
          schema:
            example: 45463597-20b7-4af7-b3b3-f5fb778203ab
            type: string
      responses:
        '200':
          description: The transcription job's metadata
          content:
            application/json:
              schema:
                example:
                  id: 45463597-20b7-4af7-b3b3-f5fb778203ab
                  request_id: G-45463597
                  version: 2
                  kind: pre-recorded
                  created_at: '2023-12-28T09:04:17.210Z'
                  status: queued
                  file:
                    id: f0dcZE10-23d8-47f0-a25d-74a6eed88721
                    filename: split_infinity.wav
                    source: >-
                      http://files.gladia.io/example/audio-transcription/split_infinity.wav
                    audio_duration: 20
                    number_of_channels: 1
                  request_params:
                    audio_url: >-
                      http://files.gladia.io/example/audio-transcription/split_infinity.wav
                    subtitles: false
                    diarization: false
                    translation: false
                    summarization: false
                    sentences: false
                    moderation: false
                    named_entity_recognition: false
                    name_consistency: false
                    custom_spelling: false
                    structured_data_extraction: false
                    chapterization: false
                    sentiment_analysis: false
                    display_mode: false
                    audio_enhancer: false
                    language_config:
                      code_switching: false
                      languages:
                        - fr
                        - en
                    accurate_words_timestamps: false
                    diarization_enhanced: false
                    punctuation_enhanced: false
                  completed_at: null
                  custom_metadata: null
                  error_code: null
                  result: null
                oneOf:
                  - $ref: '#/components/schemas/PreRecordedResponse'
                  - $ref: '#/components/schemas/StreamingResponse'
                discriminator:
                  propertyName: kind
                  mapping:
                    pre-recorded: '#/components/schemas/PreRecordedResponse'
                    live: '#/components/schemas/StreamingResponse'
        '401':
          description: You don't have the permissions to access the transcription job
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/UnauthorizedErrorResponse'
        '404':
          description: The transcription job doesn't exist or has been deleted
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/NotFoundErrorResponse'
      security:
        - x_gladia_key: []
components:
  schemas:
    PreRecordedResponse:
      type: object
      properties:
        id:
          type: string
          description: Id of the job
          format: uuid
          example: 45463597-20b7-4af7-b3b3-f5fb778203ab
        request_id:
          type: string
          description: Debug id
          example: G-45463597
        version:
          type: integer
          description: API version
          example: 2
        status:
          type: string
          description: >-
            "queued": the job has been queued. "processing": the job is being
            processed. "done": the job has been processed and the result is
            available. "error": an error occurred during the job's processing.
          enum:
            - queued
            - processing
            - done
            - error
        created_at:
          type: string
          description: Creation date
          format: date-time
          example: '2023-12-28T09:04:17.210Z'
        completed_at:
          type: string
          description: Completion date when status is "done" or "error"
          format: date-time
          example: '2023-12-28T09:04:37.210Z'
          nullable: true
        custom_metadata:
          type: object
          description: Custom metadata given in the initial request
          example: 

```

--------------------------------

### Get Pre-recorded Audio Translation Result

Source: https://docs.gladia.io/api-reference/v2/pre-recorded/get

Retrieves the translation and transcription results for a pre-recorded audio file. This endpoint is useful for getting a comprehensive analysis of the audio content, including full transcripts, detected languages, and sentence-level details.

```APIDOC
## GET /v2/pre-recorded/get

### Description
Retrieves the translation and transcription results for a pre-recorded audio file.

### Method
GET

### Endpoint
/v2/pre-recorded/get

### Parameters
#### Query Parameters
- **id** (string) - Required - The unique identifier of the pre-recorded audio job.

### Response
#### Success Response (200)
- **id** (string) - The unique identifier of the pre-recorded audio job.
- **status** (string) - The current status of the job (e.g., `processing`, `completed`, `failed`).
- **type** (string) - The type of the audio intelligence task performed (e.g., `translation`, `transcription`).
- **results** (object) - Contains the detailed results of the audio processing.
  - **TranslationResultDTO** (object) - Object containing translation-specific results.
    - **error** (AddonErrorDTO) - Contains error details if the translation failed.
    - **full_transcript** (string) - The complete transcription of the audio in text format.
    - **languages** (array) - An array of detected language codes, sorted by detection frequency.
    - **sentences** (array) - An array of sentence-level transcription and translation data, if enabled.
    - **subtitles** (array) - An array of subtitle data, if enabled.
    - **utterances** (array) - An array of transcribed speech utterances.

#### Response Example
```json
{
  "id": "a1b2c3d4-e5f6-7890-1234-567890abcdef",
  "status": "completed",
  "type": "translation",
  "results": {
    "TranslationResultDTO": {
      "error": null,
      "full_transcript": "Hello, this is a test transcript.",
      "languages": ["en"],
      "sentences": [
        {
          "start": 0.5,
          "end": 3.2,
          "text": "Hello, this is a test transcript.",
          "translation": "Bonjour, ceci est une transcription de test."
        }
      ],
      "subtitles": [
        {
          "start": 0.5,
          "end": 3.2,
          "text": "Hello, this is a test transcript."
        }
      ],
      "utterances": [
        {
          "start": 0.5,
          "end": 3.2,
          "text": "Hello, this is a test transcript."
        }
      ]
    }
  }
}
```
```

--------------------------------

### Sentiment Analysis Message

Source: https://docs.gladia.io/api-reference/v2/live/websocket

This message provides detailed information about the sentiment and emotion analysis of an utterance. It includes the original utterance text, start and end times, and a list of results, each containing sentiment, emotion, text, start, end, and channel information.

```APIDOC
## Sentiment Analysis Message

### Description
This message contains the informations about the sentiment analysis of the audio chunk.

### Payload Schema
```json
{
  "session_id": "string",
  "created_at": "string",
  "type": "sentiment_analysis",
  "error": null | object,
  "data": {
    "utterance_id": "string",
    "utterance": {
      "text": "string",
      "start": "number",
      "end": "number"
    },
    "results": [
      {
        "sentiment": "string",
        "emotion": "string",
        "text": "string",
        "start": "number",
        "end": "number",
        "channel": "integer"
      }
    ]
  }
}
```

### Example
```json
{
  "session_id": "550e8400-e29b-41d4-a716-446655440000",
  "created_at": "2025-09-19T12:34:35Z",
  "type": "sentiment_analysis",
  "error": null,
  "data": {
    "utterance_id": "utt_003",
    "utterance": {
      "text": "I love this product!",
      "start": 11,
      "end": 13.2
    },
    "results": [
      {
        "sentiment": "positive",
        "emotion": "joy",
        "text": "love this product",
        "start": 11.4,
        "end": 12.9,
        "channel": 0
      }
    ]
  }
}
```
```

--------------------------------

### SpeechStartMessage

Source: https://docs.gladia.io/api-reference/v2/live/websocket

Indicates the start of detected speech activity on a given channel.

```APIDOC
## SpeechStartMessage

### Description
Indicates the start of detected speech activity on a given channel.

### Method
SEND

### Message Type
speech_start

### Payload
- **session_id** (string) - UUID identifying the session.
- **created_at** (string) - ISO 8601 timestamp in UTC.
- **type** (string) - Must be "speech_start".
- **data** (object) - Timing information for speech start/end events.
  - **time** (number) - Time in seconds from session start when the event occurred.
  - **channel** (integer) - Audio channel index (0-based) for which the event applies.

### Example
```json
{
  "session_id": "550e8400-e29b-41d4-a716-446655440000",
  "created_at": "2025-09-19T12:34:06Z",
  "type": "speech_start",
  "data": {
    "time": 1.24,
    "channel": 0
  }
}
```
```

--------------------------------

### Initialize AssemblyAI Client (TypeScript)

Source: https://docs.gladia.io/chapters/migrations/from-assembly

Initialize the AssemblyAI client with your API key.

```typescript
import { AssemblyAI } from "assemblyai";

const assemblyClient = new AssemblyAI({
  apiKey: process.env.ASSEMBLYAI_API_KEY!,
});
```

--------------------------------

### Enable Summarization for Live Audio

Source: https://docs.gladia.io/chapters/audio-intelligence/summarization

Configure 'post_processing' and 'summarization' to true for live audio summarization. Ensure 'receive_post_processing_events' is enabled.

```json
{
  "post_processing": {
    "summarization": true,
    "summarization_config": {
      "type": "concise"
    }
  },
  "messages_config": {
    "receive_post_processing_events": true
  }
}
```

--------------------------------

### Initialize Live Transcription with Multiple Channels

Source: https://docs.gladia.io/chapters/limits-and-specifications/multiple-channels

Specify the number of channels when initializing a live transcription session. This allows Gladia to split and transcribe channels separately.

```json
{
  "channels": 2
}
```

--------------------------------

### Livekit Integration

Source: https://docs.gladia.io/llms.txt

How to integrate Gladia with LiveKit Agents.

```APIDOC
## Livekit Integration

### Description
How to integrate Gladia with LiveKit Agents.
```

--------------------------------

### Initiate Live Transcription Session (cURL)

Source: https://docs.gladia.io/chapters/live-stt/quickstart

Initiates a live transcription session using cURL. This command sends audio configuration to the Gladia API, returning a WebSocket URL for subsequent connections. Replace YOUR_GLADIA_API_KEY with your actual API key.

```bash
curl --request POST \
  --url https://api.gladia.io/v2/live \
  --header 'Content-Type: application/json' \
  --header 'x-gladia-key: YOUR_GLADIA_API_KEY' \
  --data '{ \
    "encoding": "wav/pcm", \
    "sample_rate": 16000, \
    "bit_depth": 16, \
    "channels": 1 \
    } \
  '
```

--------------------------------

### Get Live Transcription

Source: https://docs.gladia.io/api-reference/v2/transcription/get

Retrieves the status and results of a live audio transcription job.

```APIDOC
## GET /v2/transcription/{transcriptionId}

### Description
Retrieves the status and results of a live audio transcription job.

### Method
GET

### Endpoint
/v2/transcription/{transcriptionId}

### Parameters
#### Path Parameters
- **transcriptionId** (string) - Required - The ID of the transcription job to retrieve.

### Response
#### Success Response (200)
- **id** (string) - Id of the job
- **request_id** (string) - Debug id
- **version** (integer) - API version
- **status** (string) - "queued", "processing", "done", or "error"
- **created_at** (string) - Creation date
- **completed_at** (string) - Completion date when status is "done" or "error" (nullable)
- **custom_metadata** (object) - Custom metadata given in the initial request (additionalProperties: true)
- **error_code** (integer) - HTTP status code of the error if status is "error" (nullable)
- **post_session_metadata** (object) - For debugging purposes, send data that could help to identify issues
- **kind** (string) - "live"
- **file** (object) - The file data you uploaded. Can be null if status is "error" (nullable)
- **request_params** (object) - Parameters used for this live transcription. Can be null if status is "error" (nullable)
- **result** (object) - Live transcription's result when status is "done" (nullable)

### Response Example
{
  "example": {
    "id": "45463597-20b7-4af7-b3b3-f5fb778203ab",
    "request_id": "G-45463597",
    "version": 2,
    "status": "done",
    "created_at": "2023-12-28T09:04:17.210Z",
    "completed_at": "2023-12-28T09:04:37.210Z",
    "custom_metadata": {
      "user": "John Doe"
    },
    "error_code": null,
    "post_session_metadata": {},
    "kind": "live",
    "file": {},
    "request_params": {},
    "result": {}
  }
}
```
Tokens:
Raw