Documentation
¶
Overview ¶
Package foundry adapts the OpenAI Realtime provider to Microsoft Foundry. Foundry serves the identical OpenAI Realtime WebSocket protocol from the Azure-hosted v1 surface (wss://<account-host>/openai/v1/realtime?model= <deployment>), so this package wraps live/openai and only changes identity, endpoint enforcement, and the default deployment name. Auth is the shared Foundry API key sent as a Bearer token, which the v1 surface accepts.
Index ¶
Constants ¶
const DefaultRealtimeModel = openai.DefaultRealtimeModel
DefaultRealtimeModel is the default Foundry realtime deployment name. Foundry's model parameter addresses a deployment, so users who named their deployment differently override this via config.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type Provider ¶
Provider implements live.LiveProvider against a Microsoft Foundry realtime deployment by delegating the full OpenAI Realtime protocol to live/openai.
func (*Provider) Connect ¶
Connect requires the Foundry realtime endpoint (wss://<account-host>/openai/v1/realtime) in cfg.Endpoint and then dials via the embedded OpenAI Realtime implementation.
func (*Provider) SessionCapabilities ¶
func (p *Provider) SessionCapabilities() live.SessionCapabilities
Source Files