🎤 Speech Recognition and Speaker Recognition Service

Waiting to start recording...

🔄 Turn Detection Events

Enable Turn Detection: Turn Detection: Off

Waiting for turn detection events...

Active Speaker

Turn Ended

End-of-Utterance

✂️ Segments Visualization

Waiting for segment data...

Segment timeline will appear here

🎙️ Speaker Registration

Speaker Name *

🔍 Speaker Identification

👥 Speaker Management

Loading speaker list...

Realtime LLM Connection

Realtime connection: disconnected

WebSocket URL

Send response.create

Instructions / Prompt *

Response ID (optional)

Modalities (comma separated)

Model override (metadata.model)

LLM provider tag (metadata.provider)

Extra metadata (JSON)

Responses API Tester

Use this form to hit any OpenAI-compatible /v1/responses endpoint (local Henir proxy or remote provider). The stored API key is used for both X-API-Key and Authorization headers.

Endpoint URL

Model ID

Instructions (system / developer)

User prompt *

Metadata (JSON, optional)

Responses output will appear here...

🤖 LLM Events & Responses

When connected to the OpenAI-compatible Realtime endpoint, streamed events such as response.output_text.delta and conversation.item.create will appear here.

Waiting for LLM responses...

Waiting for conversation events...

Henir Speech & Speaker Recognition Service