Releasesby GlyphImproved models
Batch and streaming models Polish ( pl ) Hebrew ( he ) Batch models Macedonian ( mk ) Streaming models Estonian ( et ) Latvian ( lv ) Lithuanian ( lt ) Malay ( ms ) You do not need to change your API requests to access the improvements to existing languages; the updates are live for all users.
Also shipped Sep 3, 2026Deepgram in September 2026 →
Conversational analytics now supports questions about market basket analysis
Conversational analytics now supports questions about market basket analysis . This feature is generally available (GA).
External lineage (General availability)
External lineage is now generally available and is no longer in Preview .
Restricted Session Scope for agents (General availability)
With this release, Restricted Session Scope (RSS) is generally available. An RSS is a privilege ceiling that limits what an agent can do on behalf of a user. It intersects with RBAC and never grants privileges the user doesn't already have through their roles.
Share metric views with OpenSharing (Beta)
You can now share metric views with Databricks users in other metastores or accounts using OpenSharing. See Add metric views to a share and Read shared metric views .
Genie One enhancements
- Ask Genie button hidden when unsupported : In embedded dashboards, the Ask Genie button is now hidden when it isn't supported. See Ask Genie in embedded dashboards . Push notifications for scheduled tasks (Public Preview) : The Genie mobile app can now send push notifications for scheduled tasks to iOS devices. See Push notifications .
Web search in Genie One (Beta)
Genie One can now search the public web to answer questions that need current information, such as recent releases, third-party documentation, or news. Genie One decides when to search and cites the sources it uses, showing each as a link you can open. See Search the web .
More from DeepgramEvery Deepgram release →
Nova-3 Adds Kazakh, Plus Improved Models for Estonian, Hebrew, Latvian, Lithuanian, Macedonian, Malay, and Polish
We have added Kazakh as a new Nova-3 language and released improved Nova-3 monolingual models for seven existing languages. These updates expand language coverage and enhance transcription quality for batch and streaming workloads.
New language
- Kazakh ( kk , kk-KZ ) , available for batch and streaming. Access this language by setting model="nova-3" and language="kk" in your request.
Wider speed range for Flux TTS
Flux TTS speed now runs from 0.5 to 1.5 in 0.05 increments, up from 0.85 – 1.15 . The default stays 1.0 , and every previously accepted value still works. The wider range applies everywhere speed is accepted on Flux TTS: the /v2/speak streaming WebSocket (at connection and mid-session with Configure ), the /v2/speak batch REST endpoint, and...
Hold a function call until the user's turn is confirmed
The agent begins building a reply before speech-to-text confirms the user has finished speaking. Agent audio is held until that confirmation, but function calls have always dispatched as soon as the LLM emitted them.
defer_until_eot
Add defer_until_eot: true to any function in agent.think.functions to hold it until the turn is confirmed. If the user keeps speaking and the turn resumes, the call is discarded before it does anything.
FunctionCallCancelled
Clients now receive an explicit signal when a client-side function call they were sent is cancelled because the user started speaking again, either inside the speculative window (the turn resumed) or after the turn was confirmed (barge-in). Only calls the client already received are announced.
Flexible turn-taking control for Flux
Flux now gives you full control over how turns end. Beyond Flux's native end-of-turn detection, you can override it, suppress it entirely, or blend the two , and switch between these modes on the fly, even for a single turn, with a Configure message. Turn-taking goes from a fixed behavior to something you shape around each moment of a conversation.
Three modes of turn-taking
- Automatic (default). Flux's native end-of-turn detection decides when a turn ends. EndOfTurn events carry "trigger": "model" . Nothing changes for existing integrations. Semi-manual. Keep Flux's detection running, but override it whenever you have a definitive signal , a push-to-talk release, a DTMF tone, a "send" tap , by sending a ForceEndTurn .