Releasesby GlyphImproved models
Batch and streaming models Catalan ( ca ) Japanese ( ja ) You don't need to change your API requests to access the improvements to existing languages; the updates are live for all users. For more details and to see the full list of supported languages, visit the Models & Languages Overview page.
Also shipped Nov 12, 2026Deepgram in November 2026 →
Genie ZeroOps is available in Beta
Genie ZeroOps is now in beta, providing automated monitoring for jobs, pipelines, and tables. It detects failures and data quality issues, investigates root causes, and proposes fixes.
You can track table-level data lineage in Knowledge Catalog for BigQuery pipelines
You can track table-level data lineage in Knowledge Catalog for BigQuery pipelines. For more information, see Track data lineage . This feature is generally available (GA).
BigQuery generative AI functions now support the gemini-4-hydrogen model
BigQuery generative AI functions now support the gemini-4-hydrogen model. This model is generally available .
Amazon Aurora and RDS now support Graviton5-based R9g database instances
AWS Graviton5-based R9g database instances are now generally available for Amazon Aurora with PostgreSQL compatibility and MySQL compatibility, and for Amazon RDS for PostgreSQL, MySQL, and MariaDB.
[In preview] Public Preview: SQL performance monitoring for Azure SQL Managed Instance
The public preview of Microsoft-managed performance monitoring for Azure SQL Managed Instance is now available. Customers monitoring business-critical SQL Managed Instance workloads often rely on custom scripts and disconnected monitoring tools. This requ
Gemini Enterprise: Gemini 4 Hydrogen available in web app, mobile app, Workflow Builder, and AI developer tools
Gemini 4 Hydrogen is generally available (GA) for the Gemini Enterprise web app, mobile app, Workflow Builder, and AI developer tools. Location availability : Supported in global , us , and eu multi-regions, with data residency and ML processing support in us and eu . For more information, see Data residency for Gemini Enterprise .
More from DeepgramEvery Deepgram release →
Improved Nova-3 Models for Catalan and Japanese
We've released improved Nova-3 monolingual models for Catalan and Japanese. These updates enhance transcription quality for batch and streaming workloads.
Improved Nova-3 Models for Arabic, Armenian, Bengali, Bulgarian, Danish, English (Australia, New Zealand), Greek, Gujarati, Hebrew, Nepali, Slovak, Tagalog, and Urdu
We've released improved Nova-3 monolingual models for 13 existing languages. These updates enhance transcription quality for batch and streaming workloads.
Improved Model for English (Australia, New Zealand)
We've added a Nova-3 English model for Australian and New Zealand English, available for batch and streaming alongside the general Nova-3 English model. Requests to model=nova-3 with language=en-AU or language=en-NZ use the new model automatically; all other English codes continue to use the general model. No changes to your API requests are needed.
Container Images (release 261110)
Deepgram released updated container images for its self-hosted API, engine, license proxy, and billing services. These images require a minimum NVIDIA driver version of 580.
FIPS 140-3 Images
Deepgram now provides FIPS 140-3 compliant container images for all self-hosted services. These images use OpenSSL's validated cryptographic provider and can be enabled via configuration.
This Release Contains The Following Changes
Deepgram updated its API to enforce stricter validation for language hints on multilingual models and deprecated the dg-token header for callback verification in favor of Basic Auth.
Verify Callback Requests with Basic Auth or Extra Metadata
The dg-token header is no longer supported for verifying callback requests. Users must switch to Basic Auth or use a secret passed via Extra Metadata.
Speech-to-Text ( /v1/listen )
Deepgram added Basic Auth support via callback URLs and secret passing via extra metadata to its Speech-to-Text endpoint, enabling server-side callback request verification.