- Featuredevelopers.deepgram.com
Deepgram adds Extra Metadata feature to API responses
Deepgram introduced Extra Metadata, allowing users to attach arbitrary key-value pairs to API requests that are returned in responses for downstream processing. The feature supports up to 2048 charact…
- Featuredevelopers.deepgram.com
Deepgram launches MCP servers and skills marketplace for AI coding tools
Deepgram introduced a CLI with a built-in MCP server, a docs MCP server for querying documentation, and a skills marketplace with structured reference material for AI-powered developer tools. These to…
- developers.deepgram.com
Deepgram launches streaming text-to-speech WebSocket API with SDKs
Deepgram introduced a new WebSocket-based text-to-speech (TTS) API that converts streaming text into speech in real time. The feature supports multiple SDKs (JavaScript, Python, Go, C#, Java) and uses…
- Featuredevelopers.deepgram.com
Deepgram expands Aura-2 TTS voices with 30+ English options and broader language support
Deepgram’s Aura-2 text-to-speech API now offers over 30 English voices across American, British, Australian, Irish, and Filipino accents, alongside expanded language support including Spanish, German,…
- developers.deepgram.com
Deepgram launches Think-Fast and phoneLLM mini LLMs for voice agents
Deepgram introduced two new LLMs—Think-Fast for instruction-following voice conversations and phoneLLM mini for speed-optimized pipelines—both available during Early Access. Think-Fast supports cascad…
- developers.deepgram.com
Deepgram launches Early Access LLM endpoint /v1/think with Think-Fast model
Deepgram introduced a new LLM endpoint (/v1/think) and model (think-fast) for Early Access customers, available only to contract holders. The endpoint supports stateless chat completions with system/u…
- Featuredevelopers.deepgram.com
Deepgram launches Early Access LLMs for voice workloads with Think-Fast and phoneLLM mini
Deepgram introduced two Early Access LLMs—Think-Fast and phoneLLM mini—designed for voice agent pipelines. Think-Fast supports cascaded agent workflows with tool calling, while phoneLLM mini prioritiz…
- Launchdeepgram.com
Deepgram launches self-hosted voice AI for cloud and on-premises deployment
Deepgram introduced a self-hosted version of its voice AI platform, enabling deployment on customer infrastructure for enhanced performance and privacy. The offering supports real-time inference with …
- Launchdeepgram.com
Deepgram launches Voice Agent Platform with Think-Live/Think-Fast LLMs and Real-Time Context Standard
Deepgram introduced the Real-Time Context Standard (RCS) for voice agents, two specialized LLMs (Think-Live for conversation management and Think-Fast for task execution), and a new Voice Agent Platfo…
- deepgram.com
Deepgram acquires Daily and Pipecat to launch first open, production-ready voice AI stack
Deepgram announced the acquisition of Daily, creators of the open-source Pipecat voice agent framework, to integrate Pipecat’s orchestration layer with Deepgram’s speech models and Daily’s global real…
- Featuredevelopers.deepgram.com
Deepgram deprecates dg-token for callback authentication
Deepgram is retiring the `dg-token` header for verifying callback requests across its Speech-to-Speech, Text-to-Speech, and Text Intelligence APIs. Developers must migrate to Basic Auth or use a secre…
- developers.deepgram.com
Deepgram Nova-3 adds multilingual transcription with language hints
Deepgram’s Nova-3 model now supports multilingual transcription for 55 languages, including mid-conversation language switching, via the `language=multi` parameter. Users can optionally bias detection…
- Featuredevelopers.deepgram.com
Deepgram Nova-3 multilingual model expands to 55 languages with language hinting
Deepgram’s Nova-3 multilingual model now supports 55 languages (up from 10) and introduces language hinting to bias detection toward expected languages. Code-switching and per-word language tags remai…
- deepgram.com
Resources Hub
Deepgram announced the acquisition of Daily, the company behind Pipecat, to create the first complete, production-ready open stack for conversational voice AI. This move integrates Daily's real-time c…

- Featuredevelopers.deepgram.com
Deepgram launches EU, AU, and India regional endpoints for data residency
Deepgram introduced regional endpoints in the EU (`api.eu.deepgram.com`), Australia (`api.au.deepgram.com`), and India (`api.in.deepgram.com`) to ensure data processing occurs within the specified geo…
- Featuredevelopers.deepgram.com
Deepgram Self-Hosted 261020 release: Flux STT runtime upgrade, Nova-3 keyterm updates, and FIPS 140-3 support
Deepgram’s October 20, 2026 self-hosted release (261020) introduces a new Flux STT Engine runtime requiring updated .dgv2 model files, mid-stream keyterm updates for Nova-3 models, and always-availabl…
- Featuredevelopers.deepgram.com
Deepgram Nova-3 Configure updates now return single success/failure messages
Deepgram’s Nova-3 streaming API now responds to every Configure message on a /v1/listen stream with exactly one ConfigureSuccess or ConfigureFailure message. ConfigureSuccess includes the active keyte…
- deepgram.com
Deepgram Flux STT powers Anam’s avatar turn-taking with 50% fewer barge-ins
Deepgram’s Flux STT, designed for conversational turn-taking, was adopted by Anam to power its photorealistic interactive avatars. The integration reduced unwanted avatar interruptions by ~50% with no…
- Integrationdeepgram.com
Deepgram Flux STT powers Anam’s photorealistic avatar turn-taking
Deepgram’s Flux STT model was adopted by Anam to solve real-time turn-taking challenges in unscripted avatar conversations, reducing unwanted barge-ins by ~50% with no latency penalty. Anam deployed F…
- Featuredevelopers.deepgram.com
Deepgram SDKs, CLI, and Skills receive mid-stream configuration, TTS playback, and regional endpoint support
Deepgram released updates across multiple SDKs (Python v7.13.0, JavaScript v5.15.0, Java v0.11.1, Go v3.9.0, .NET v7.1.3, Rust v0.12.0/v0.13.0), CLI v0.3.3, and Skills v1.7.0. Key additions include mi…
- Integrationdeepgram.com
Deepgram integrates with Jev classifier in five open-source voice agent projects
Deepgram published five open-source projects demonstrating integrations with Jev, a text-based classifier, to power voice-first applications. Projects include a Star Wars droid voice assistant, a no-L…
- Contentdeepgram.com
Deepgram highlights five open-source projects using Jev + Deepgram for voice decision-making
Deepgram published a case-study-style blog post featuring five open-source projects that combine Deepgram’s speech-to-text (STT) with Jev’s classifier to handle voice-based decisions without generativ…
- Featuredevelopers.deepgram.com
Deepgram corrects keyterm update behavior in Nova-3 streaming API
Deepgram’s changelog fixes a documentation error about over-limit keyterm updates in its Nova-3 streaming API. Previously, exceeding the 500-token limit was said to return an error, but it now stops t…
- Contentdeepgram.com
Why I've Stayed with the Same Startup for 4 Years
A Deepgram employee details their four-year tenure, highlighting the company’s culture, product ambition, and recent milestones like a $130M Series C and launches such as Nova-3 Pharmab. The post emph…

- Featuredevelopers.deepgram.com
Deepgram improves Nova-3 Medical English models with fewer word and entity errors
Deepgram released enhanced Nova-3 Medical English models for both batch and streaming transcription, reducing word errors and medical entity inaccuracies. The updates are live for all users without AP…
- deepgram.com
Introducing Nova-3 Medical Multilingual: Medical Speech Recognition in 10 Languages
Deepgram extended its Nova-3 Medical speech recognition model to support 10 languages (Dutch, English, French, German, Hindi, Italian, Japanese, Portuguese, Russian, Spanish) with medical specializati…

- deepgram.com
Deepgram Nova-3 Medical Multilingual: 10-language medical speech recognition with 25-40% fewer errors
Deepgram launched Nova-3 Medical Multilingual, extending its medical speech-to-text model to 10 languages (Dutch, English, French, German, Hindi, Italian, Japanese, Portuguese, Russian, Spanish) with …
- Contentdeepgram.com
AI voice agent to human handoffs: How to warm transfer calls without losing context
Deepgram published a technical guide on implementing warm transfers for AI voice agents, highlighting how structured context preservation (identity, intent, actions) reduces caller frustration and spe…
- Featuredevelopers.deepgram.com
Deepgram Nova-3 Medical adds multilingual support for 10 languages
Deepgram expanded its Nova-3 Medical model to support multilingual transcription with a single `language=multi` parameter, enabling real-time and batch transcription for medical audio where speakers s…
- Featuredeepgram.com
Deepgram Voice Agent API adds warm transfer with context preservation
Deepgram introduced guidance for warm transfers in its Voice Agent API, enabling AI-to-human handoffs without losing context. The update includes escalation triggers, structured context payloads (iden…
- developers.deepgram.com
Deepgram Nova-3 models improved for Catalan, French, Indonesian, and Vietnamese
Deepgram released updated Nova-3 monolingual models for Catalan, French, Indonesian, and Vietnamese, enhancing batch transcription for Vietnamese and Catalan and streaming for French and Indonesian. I…
- Integrationdevelopers.deepgram.com
Amazon Connect and Deepgram
Deepgram now offers an integration with Amazon Connect, allowing organizations to replace AWS Transcribe and Polly with Deepgram’s STT and TTS models without altering existing Connect flows. The integ…
- Featuredevelopers.deepgram.com
Deepgram launches Flux model for voice agents with smart turn detection
Deepgram introduced Flux, a new model designed for voice agents with first-of-its-kind model-integrated end-of-turn detection, configurable turn-taking dynamics, and ultra-low latency (~260ms). Flux s…
- Featuredevelopers.deepgram.com
Deepgram Flux adds model-integrated end-of-turn detection and configurable voice agent patterns
Deepgram’s Flux model now includes first-of-its-kind model-integrated end-of-turn detection, configurable turn-taking dynamics, and ultra-low latency optimized for voice agent pipelines, maintaining N…
- Case Studydeepgram.com
Sharpen elevates the contact center customer experience with Deepgram
Sharpen Technologies replaced its outdated ASR solution with Deepgram’s automatic speech recognition platform, achieving significantly higher transcription accuracy and speed. The upgrade enabled Shar…
- Integrationdeepgram.com
Deepgram creates increased insights for CallTrackingMetrics customers
Deepgram’s AI Speech Platform improved CallTrackingMetrics’ transcription accuracy to over 90%, enabling better analytics for marketing attribution and contact center operations. The integration lever…
- Case Studydeepgram.com
Revenue.io boosts sales analysis with Deepgram's customized speech models
Deepgram helped Revenue.io improve transcription accuracy for sales conversations by training customized speech models using 20+ hours of audio. This tailored approach outperformed big-tech ASR provid…
- Featuredevelopers.deepgram.com
Deepgram CLI adds new text-to-speech commands and model options
Deepgram expanded its developer CLI with new text-to-speech (TTS) commands via the `dg speak` utility, supporting multiple models (Flux and Aura), output formats (WAV/MP3), and streaming playback. The…
- Featuredevelopers.deepgram.com
Deepgram CLI adds text intelligence features: sentiment, topics, summarization, and intents
Deepgram expanded its CLI tool with new text intelligence commands for sentiment analysis, topic detection, summarization, and intent recognition. Users can now analyze files, piped input, or transcri…
- Integrationdeepgram.com
Democratizing Speech Analytics: Deepgram + Callbi
Deepgram integrated its high-accuracy speech recognition API with Callbi’s cloud-based contact center analytics platform, enabling cost-effective speech analytics for contact centers of all sizes. Cal…
- Contentdeepgram.com
Compliance Monitoring for Call Centers
Deepgram published a technical tutorial showing how its speech-to-text API can integrate into call center compliance workflows, including real-time transcription, speaker diarization, phrase search, a…
- Launchdeepgram.com
Introducing Speech Summarization Powered by Domain-Specific Language Models
Deepgram introduced a new speech summarization API powered by domain-specific language models (DSLMs) tailored for contact centers and sales enablement. The model automates call summarization to reduc…
- Launchdeepgram.com
Introducing Nova-2: The Fastest, Most Accurate Speech-to-Text API
Deepgram introduced Nova-2, a new speech-to-text model delivering an 18.4% reduction in word error rate (WER) over Nova-1 and outperforming competitors by 30% in accuracy benchmarks. The model support…

- Launchdeepgram.com
Introducing Deepgram Aura: Lightning Fast Text-to-Speech for Voice AI Agents
Deepgram officially announced the general availability of Aura, its Text-to-Speech (TTS) API, expanding its Voice AI platform to include both speech recognition and voice generation. Aura delivers sub…

- Integrationdocs.aws.amazon.com
AWS Connect adds third-party TTS support including Deepgram
AWS Connect now supports configuring third-party text-to-speech (TTS) providers, including Deepgram, via a new contact flow integration. Users can select Deepgram as a voice provider, specify model/vo…
- Integrationdocs.aws.amazon.com
Deepgram integrated as third-party STT provider for AWS Connect
AWS Connect now supports Deepgram as a third-party speech-to-text (STT) provider in its bot configuration workflow. Users can select Deepgram in the Voice provider dropdown, enter a model ID, and link…
- developers.deepgram.com
Deepgram introduces Keyterm Prompting to boost transcription accuracy for industry terms
Deepgram launched Keyterm Prompting, a feature that instantly improves recognition accuracy for up to 100 important terms, product names, and jargon in transcription tasks. Available for Nova-3 and Fl…
- developers.deepgram.com
Deepgram adds real-time stream configuration via 'Configure' message for Nova-3 models
Deepgram introduced a 'Configure' message for its `/v1/listen` WebSocket streaming API, allowing dynamic adjustments to keyterms and formatting features (e.g., numerals) mid-stream without dropping au…
- Featuredevelopers.deepgram.com
Deepgram adds mid-stream keyterms updates for Nova-3 models
Deepgram introduced the ability to dynamically swap keyterms mid-stream for Nova-3 models during active /v1/listen sessions using a Configure message. The update replaces the entire keyterms list, enf…
- Featuredevelopers.deepgram.com
Deepgram Self-Hosted 261001: Flux TTS watermarking, Whisper removal, FIPS updates
Deepgram’s October 2026 Self-Hosted release (261001) adds mandatory Flux TTS audio watermarking, removes Whisper model support, and introduces FIPS 140-3 compliant images with updated metrics and cert…
Track Deepgram on autopilot
- · Weekly AI brief: narrative summary of what shipped, every Monday 9 AM
- · Email or Slack alerts, or chat with the archive in your dashboard
- · Add Deepgram + up to 2 more competitors free, no credit card