Improved models
Improved batch and streaming models for German and Hindi are now live for all users. No changes to API requests are required to access these improvements.
Also shipped Nov 17, 2026Deepgram in November 2026 →
Amazon EMR 8.2 is now available
AWS released Amazon EMR 8.2, upgrading to Apache Spark 4.1.3, Apache Iceberg 1.11.0, and Delta Lake 4.2.0 while enabling concurrent Delta Lake writes on Amazon S3.
Introducing the Apache Spark Optimization Agent for Amazon EMR Serverless
AWS introduced the Apache Spark optimization agent for Amazon EMR Serverless to evaluate completed Spark jobs and output automated performance and cost tuning recommendations.
Amazon Bedrock AgentCore Memory now provides full session summaries
Amazon Bedrock AgentCore Memory now automatically creates and maintains a full session summary record for sessions configured with a summary strategy.
[Launched] Generally Available: Intel and AMD v6 with temp disk for Azure Database for PostgreSQL
Microsoft announced general availability of Intel and AMD v6 compute options with local temporary disk for Azure Database for PostgreSQL Flexible Server.
Amazon SageMaker now supports AI data preparation capabilities to build high-quality datasets for model development
Amazon SageMaker introduced AI data preparation features, including synthetic data generation, automated multi-task labeling, human review queues, and quality evaluation.
BigQuery AI.IF and AI.CLASSIFY optimized mode is generally available
Google Cloud made optimized mode generally available for BigQuery managed AI functions AI.IF and AI.CLASSIFY to lower query latency and token usage.
More from DeepgramEvery Deepgram release →
Improved Nova-3 Models for German and Hindi
Deepgram has released improved Nova-3 monolingual models for German and Hindi, enhancing transcription quality for both batch and streaming workloads.
Nova-3 Adds 36 Languages: Abkhazian, Adyghe, Akan, Albanian, Amharic, Azerbaijani, Basque, Bodo, Central Kurdish, Divehi, Eastern Mari, Galician, Hausa, Icelandic, Kabardian, Kabyle, Kalenjin, Kashmir
Deepgram added 36 new monolingual languages to the Nova-3 speech recognition model, expanding language coverage across batch and streaming workloads.
New languages
Deepgram added 36 new languages to Nova-3 for batch and streaming requests, accessible by setting model="nova-3" and the appropriate language code. Self-hosted customers can request updated model artifacts for download.
Deepgram on AWS Marketplace with x402
Deepgram speech-to-text and text-to-speech models are now available via AWS Marketplace, allowing usage-based billing through AWS accounts without requiring a separate Deepgram API key.
Deepgram CLI 0.5.0: dg speak gets --out-file, dg init needs --yes outside a terminal
Deepgram CLI v0.5.0 Deepgram CLI v0.5.0 , released 13 November, changes how dg speak saves audio, how dg init confirms setup outside a terminal, and what dg listen -o prints. dg speak adds --out-file , which refuses to overwrite an existing file unless you also pass --force .
Improved Nova-3 Models for Catalan and Japanese
Updated Nova-3 monolingual models for Catalan and Japanese are now available, providing improved transcription quality for both batch and streaming workloads.
Improved models
Improved Nova-3 models for Catalan and Japanese are live for all users. Self-hosted customers can request the updated models from their account representative.
Improved Nova-3 Models for Arabic, Armenian, Bengali, Bulgarian, Danish, English (Australia, New Zealand), Greek, Gujarati, Hebrew, Nepali, Slovak, Tagalog, and Urdu
Improved Nova-3 monolingual models have been released for 13 languages, including Arabic, Bengali, and Danish, to enhance transcription quality for batch and streaming workloads.