> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mascot.bot/llms.txt
> Use this file to discover all available pages before exploring further.
# Gemini Live API Proxy
## AsyncAPI
````yaml api-reference/asyncapi.yaml geminiProxy
id: geminiProxy
title: Gemini Live API Proxy
description: >
Connect to Google Gemini Live API with automatic viseme generation.
Audio responses include a `mascotVisemes` field with synchronized viseme data
for facial animation.
Due to a Mintlify limitation, received messages may display as `{}` in the
playground. The actual response data is being received correctly — you can
inspect it in your browser's DevTools under the Network tab → WS → Messages
(UTF-8).
servers:
- id: local
protocol: ws
host: localhost:3333
bindings: []
variables: []
- id: production
protocol: wss
host: api.mascot.bot
bindings: []
variables: []
address: >-
/v1/conversation/gemini?key={geminiApiKey}&mascot_key={mascotApiKey}&viseme_model={visemeModel}
parameters:
- id: geminiApiKey
jsonSchema:
type: string
description: Your Google Gemini API key
default: YOUR_GEMINI_API_KEY
description: Your Google Gemini API key
type: string
required: true
deprecated: false
- id: mascotApiKey
jsonSchema:
type: string
description: Your Mascot Bot API key
default: YOUR_MASCOT_API_KEY
description: Your Mascot Bot API key
type: string
required: true
deprecated: false
- id: visemeModel
jsonSchema:
type: string
description: >
Viseme model to use for lip sync prediction. Different models are
optimized for different languages.
Available: `default` (English), `indonesian` (Bahasa Indonesia).
default: default
description: >
Viseme model to use for lip sync prediction. Different models are
optimized for different languages.
Available: `default` (English), `indonesian` (Bahasa Indonesia).
type: string
required: true
deprecated: false
bindings: []
operations:
- &ref_1
id: sendSetup
title: Send setup
type: receive
messages:
- &ref_4
id: SetupMessage
payload:
- name: Setup Message
description: Initialize the Gemini session with model and configuration
type: object
properties:
- name: setup
type: object
description: Session configuration
required: true
properties:
- name: model
type: string
description: Gemini model to use
required: false
- name: generationConfig
type: object
description: Generation settings
required: false
headers: []
jsonPayloadSchema:
type: object
properties:
setup:
type: object
description: Session configuration
properties:
model:
type: string
description: Gemini model to use
default: models/gemini-2.0-flash-exp
x-parser-schema-id:
generationConfig:
type: object
description: Generation settings
default:
responseModalities:
- AUDIO
speechConfig:
voiceConfig:
prebuiltVoiceConfig:
voiceName: Aoede
x-parser-schema-id:
x-parser-schema-id:
required:
- setup
x-parser-schema-id:
title: Setup Message
description: Initialize the Gemini session with model and configuration
example: No examples found
bindings: []
extensions:
- id: x-parser-unique-object-id
value: SetupMessage
- id: x-parser-message-name
value: SetupMessage
bindings: []
extensions: &ref_0
- id: x-parser-unique-object-id
value: geminiProxy
- &ref_2
id: sendTextMessage
title: Send text message
type: receive
messages:
- &ref_5
id: TextMessage
payload:
- name: Text Message
description: Send a text message to Gemini
type: object
properties:
- name: clientContent
type: object
description: Message content
required: true
properties:
- name: turns
type: array
description: Conversation turns
required: false
- name: turnComplete
type: boolean
description: Whether this completes the turn
required: false
headers: []
jsonPayloadSchema:
type: object
properties:
clientContent:
type: object
description: Message content
properties:
turns:
type: array
description: Conversation turns
default:
- role: user
parts:
- text: hello
x-parser-schema-id:
turnComplete:
type: boolean
description: Whether this completes the turn
default: true
x-parser-schema-id:
x-parser-schema-id:
required:
- clientContent
x-parser-schema-id:
title: Text Message
description: Send a text message to Gemini
example: No examples found
bindings: []
extensions:
- id: x-parser-unique-object-id
value: TextMessage
- id: x-parser-message-name
value: TextMessage
bindings: []
extensions: *ref_0
- &ref_3
id: receiveAudioResponse
title: Receive audio response
type: send
messages:
- &ref_6
id: AudioResponse
contentType: application/json
payload:
- name: Audio Response
description: Audio response from Gemini with viseme data
type: object
properties:
- name: serverContent
type: object
description: Gemini server response containing audio
required: false
properties:
- name: modelTurn
type: object
required: false
properties:
- name: parts
type: array
required: false
- name: mascotVisemes
type: object
description: Synchronized viseme data for lip sync
required: false
properties:
- name: visemes
type: array
description: Array of viseme timing data
required: false
- name: audio_event_id
type: integer
description: Unique identifier for audio correlation
required: false
- name: subchunk_id
type: integer
description: Current chunk number
required: false
- name: total_chunks
type: integer
description: Total chunks in segment
required: false
- name: audio_duration
type: number
description: Audio duration in milliseconds
required: false
headers: []
jsonPayloadSchema:
type: object
properties:
serverContent:
type: object
description: Gemini server response containing audio
properties:
modelTurn:
type: object
properties:
parts:
type: array
items:
type: object
properties:
inlineData:
type: object
properties:
mimeType:
type: string
description: Audio MIME type
example: audio/pcm;rate=24000
x-parser-schema-id:
data:
type: string
description: Base64-encoded PCM audio data
x-parser-schema-id:
x-parser-schema-id:
x-parser-schema-id:
x-parser-schema-id:
x-parser-schema-id:
x-parser-schema-id:
mascotVisemes:
type: object
description: Synchronized viseme data for lip sync
properties:
visemes:
type: array
description: Array of viseme timing data
items:
type: object
properties:
visemeId:
type: integer
description: Viseme ID (0-21, Microsoft SAPI mapping)
x-parser-schema-id:
offset:
type: number
description: Timing offset in milliseconds
x-parser-schema-id:
x-parser-schema-id:
x-parser-schema-id:
audio_event_id:
type: integer
description: Unique identifier for audio correlation
x-parser-schema-id:
subchunk_id:
type: integer
description: Current chunk number
x-parser-schema-id:
total_chunks:
type: integer
description: Total chunks in segment
x-parser-schema-id:
audio_duration:
type: number
description: Audio duration in milliseconds
x-parser-schema-id:
x-parser-schema-id:
x-parser-schema-id:
title: Audio Response
description: Audio response from Gemini with viseme data
example: |-
{
"serverContent": {
"modelTurn": {
"parts": [
{
"inlineData": {
"mimeType": "audio/pcm;rate=24000",
"data": "CAE1/kP7f/rU/E4A0wJwBK4EAgTRA8EDawTFBLUERgWMBsoIngqnDQ..."
}
}
]
}
},
"mascotVisemes": {
"visemes": [
{
"visemeId": 4,
"offset": 12
},
{
"visemeId": 18,
"offset": 108
},
{
"visemeId": 20,
"offset": 121
},
{
"visemeId": 6,
"offset": 204
}
],
"audio_event_id": 3001,
"subchunk_id": 1,
"total_chunks": 1,
"audio_duration": 720
}
}
bindings: []
extensions:
- id: x-parser-unique-object-id
value: AudioResponse
- id: x-parser-message-name
value: AudioResponse
bindings: []
extensions: *ref_0
sendOperations:
- *ref_1
- *ref_2
receiveOperations:
- *ref_3
sendMessages:
- *ref_4
- *ref_5
receiveMessages:
- *ref_6
extensions:
- id: x-parser-unique-object-id
value: geminiProxy
securitySchemes: []
````