Create a web voice session token

POST

/api/web-access/token

Starts a browser-based voice session for an agent and returns short-lived credentials for the realtime media connection. Access is gated by the agent's configured allowed embed domains: the request Origin/Referer must match one of those domains (platform-internal origins are always permitted). Tenant quota and reseller package/pool limits are enforced before a session is created.

Request

Authorization

API Key

Add parameter in header

X-API-Key

Example:

X-API-Key: ********************

Header Params

Body Params application/jsonRequired

Examples

Responses

🟢200Success

application/json

Session created. Returns the connection token, realtime media URL, room name, basic agent info, and optional avatar streaming details.

Bodyapplication/json

🟠400

🟠402

🟠403

🟠404

🔴500

Request Example

Shell

JavaScript

Java

Swift

curl --location 'https://api.snakk.ai/api/web-access/token' \
--header 'Origin;' \
--header 'Referer;' \
--header 'X-API-Key: <api-key>' \
--header 'Content-Type: application/json' \
--data '{
    "agent_id": "11111111-1111-1111-1111-111111111111"
}'

Response Example

200 - Success

{
    "token": "eyJhbGciOiJI...",
    "url": "wss://media.example.com",
    "room_name": "call-web-a1b2c3d4",
    "agent": {
        "name": "Example AS Assistant",
        "voice": "cedar"
    },
    "avatar": {
        "enabled": false
    }
}

Modified at 2026-07-08 19:01:40

Upgrade subscription (deprecated)

Generate an agent prompt from a description