NVIDIA releases multilingual transcription model
Hugging Face: NVIDIA Nemotron 3.5 Transcribe and Granary v2: Intelligent Multilingual Transcription You Can Prompt and Adapt
Customizable transcription tools could help organizations reduce manual annotation costs while handling multiple languages simultaneously.
What happened
NVIDIA released Nemotron 3.5 Transcribe, an open-weight AI model that converts recorded speech to text while identifying speakers and translating content across multiple languages. The model supports transcription in 40 languages, translation in 31 languages, and speaker identification in 13 languages, and can process recordings up to an hour long while identifying up to eight speakers. It comes with Granary v2, an open data pipeline that lets developers customize outputs for domain-specific terminology and accents.
What each outlet adds
- Hugging Face ↗: The model ranks first among open-weight models on multiple leaderboards for both transcription accuracy and speaker attribution.
- TechCrunch ↗: The model works only on recordings, not live audio, leaving real-time transcription to rivals like ElevenLabs, and runs on a single chip across various hardware platforms.
- SiliconANGLE ↗: The model can handle 100 hours of audio and its accuracy improves significantly when fine-tuned for specialized fields like medical speech recognition.
Why it matters
Customizable transcription tools could help organizations reduce manual annotation costs while handling multiple languages simultaneously. The release occurs during an active period for voice AI, with other companies including Mistral, ElevenLabs, and Salesforce also launching voice models recently.
Get smart on it
NVIDIA Nemotron 3.5 Transcribe is an open-weight model that converts recorded speech into text while also identifying speakers, translating content, and answering questions about audio. The model supports transcription in 40 languages, translation in 31 languages, and speaker identification in 13 languages, and developers can guide its output using natural language prompts to handle domain-specific terminology and accents. This matters because a transcript alone often leaves unclear who said what, how to translate it, or how to format specialized terms, which different applications like multilingual calls, meetings, and medical dictation require in different ways. The model ranks first among open-weight models on multiple leaderboards for both transcription accuracy and speaker attribution, and comes with Granary v2, an open data pipeline and fine-tuning recipes that allow developers to further adapt the model to their specific needs.
Relevant courses
Want to go deeper than the news? Explore live, cohort-based AI courses taught by practitioners.
Browse AI courses on MavenRelated news
Introducing GPT-6.1 Terra
Bringing fast and capable computer use to everyone
Also reported by SiliconANGLE, MarkTechPost and The Decoder
Genie 4: A new generation of world simulation
Genie 4: our new world model unlocking unprecedented realism in real-time simulation.
Also reported by TechCrunch, Ars Technica, SiliconANGLE and MarkTechPost
Google’s next Genie AI model generates more realistic experiences
Google announced today its Genie 4 world model for generating interactive experiences using AI. The company says the model has four main updates: It allows for more realistic worlds, lets users update an experience in real time with a natural language prompt, can generate audio, and uses Gemini 3.8 Flash under the hood to help with its simulations. It also can create experiences that are about two minutes long, up from one minute for the previous Genie 3 model. Google is pitchi
Moonshot AI Releases Kimi K3.1 Pro: A 2.8T Total/104B Active Params (MoE) Model with 1M Context
Moonshot AI has just released Kimi K3.1 Pro, the first model in its K3.1 series. Kimi K3.1 Pro is a 3T-class open-weight* model with a 1-million-token context window. Moonshot team positions it against the strongest proprietary models on coding, agentic and knowledge benchmarks. For developers, it is cheaper API pricing than K3 and open weights yet not available. *open weights are expected to release in November 2026. TL;DR Size: 2.8T total, 104B active parameters (MoE), 1M-token context
Also reported by SiliconANGLE and The Decoder
Build specialized multimodal models with LFM3
Organizations can now create custom AI systems combining text, images, and video without building from scratch, lowering barriers to deployment.
Also reported by MarkTechPost
Mirendil Launches Fern 0.1, Its First Model, With 10M-Token Context
Mirendil has released Fern, its first AI model. Mirendil Fern is a from-scratch coding model with a context window of up to 10M tokens. Mirendil explains how that equals about 23 MB of code, or roughly half a million lines. That is a codebase on the scale of Megatron-LM, held in context at once. Fern 0.1 is the first version, built from Mirendil-Base after a short round of post-training. For developers, its like, you load a whole repo, a full training log or days of agent traces at once. TL;