Bethann Noble

After @Pinterest @Airbnb @NotionHQ @cursor_ai, today it’s @eoghan @intercom publicly sharing that they’re finding it better, cheaper, faster to use and train open models themselves rather than use APIs for many tasks. And hundreds of other companies are doing the same without sharing. Ultimately, I believe the majority of AI workflows will be in-house based on open-source (vs API). It took much more time than we anticipated but it’s happening now!

🔊Introducing Voxtral TTS: our new frontier open-weight model for natural, expressive, and ultra-fast text-to-speech 🎭Realistic, emotionally expressive speech. 🌍Supports 9 languages and accurately captures diverse dialects. ⚡Very low latency for time-to-first-audio. 🔄Easily adaptable to new voices

While Silicon Valley obsesses over AGI, @MistralAI is betting that big enterprises and sovereign nations will want to own, not rent, their intelligence My conversation with co-founder & CTO Timothee Lacroix (@tlacroix6), for his first US podcast ever Was reminded during this conversation that Mistral is barely 2.5 years old - remarkable 00:00 — Intro 01:27 — Mistral vs. The World: From Research Lab to Sovereign Power 03:48 — Inside Mistral Compute: Building an 18,000 GPU Cluster 08:42 — The Trillion-Dollar Question: Competing Without a Big Tech Parent 10:37 — The Reality of Enterprise AI: Escaping "POC Purgatory" 15:06 — Why Mistral Hires Forward Deployed Engineers (FDEs) 16:57 — The Contrarian Take: Why "Agents" are just "Workflows" 19:35 — Trust & Autonomy: The Truth About Agent Reliability 21:26 — The Missing Stack: Governance and Versioning for AI 26:24 — When Will AI Actually Work? (The 2026 Timeline) 30:33 — Beyond Chat: The "Banger" Sovereign Use Cases 35:46 — Mistral 3 Architecture: Mixture of Experts vs. Dense 43:12 — Synthetic Data & The Post-Training Bottleneck 45:12 — Reasoning Models: Why "Thinking" is Just Tool Use 46:22 — Launching DevStral 2 and the Vibe CLI 50:49 — Engineering Lessons: How to Build Frontier AI Efficiently 56:08 — Are Enterprises Ready for AGI? & The Future of Intelligence

Introducing Mistral OCR 3, a new frontier in document intelligence! 🧵👇

Don't miss the chance to see NVIDIA Nemotron, open and efficient multimodal models for agentic AI, in action. 👇 🔹 How To Build AI Agents With Open Models 🔹 A First-Of-Its-Kind Integrated Stack for Operational AI and Specialized AI Agents 🔹 Build a Graph RAG Powered by NVIDIA Nemotron LLMs and more ➡️ nvda.ws/4q19V8G at #NVIDIAGTC Washington D.C. (available for conference pass holders only)

@capydotai Genius branding

Introducing Capy - the multiplayer coding agent that ships itself

🌟 NIM 1.4 introduces significant improvements in kernel efficiency, runtime heuristics, and memory allocation, enabling up to 2.4x faster inferencing for businesses relying on quick responses and high throughput in generative AI applications. bit.ly/4hOYQUe

NVIDIA silently released the Llama 3.1 70B Instruct Nemotron model which is topping leaderboards. You can try it here: build.nvidia.com/nvidia/llama-3… I sat down with one of the researchers to learn how it performed so well and why they built this 🤙 Anything else I should ask him?

🙌 Develop innovative #LLM applications and grow your expertise through resources tailored for developers at all skill levels. 👀 Join our developer community and participate in our latest NVIDIA and LlamaIndex #DevContest ✅ ✨ bit.ly/4dU3MVk

ICYMI: Discover how developers are enhancing their #OpenUSD workflows with NVIDIA NIM Microservices. See how USD Code, USD Search, and USD Validate are transforming 3D environment creation using #generative AI copilots & agents. bit.ly/4g6orXF

📣 Announcing NVIDIA NIM Agent Blueprints, reference applications for every enterprise to make their own #AI. ➡️ nvda.ws/474d3YA These equip millions of developers with reference applications for building and deploying #generativeAI use cases. bit.ly/3yWVbCg

🌏 NVIDIA is introducing sovereign AI NIM microservices for Japan and Taiwan, enhancing #generativeAI with four new models to speed up the deployment of regional AI applications. bit.ly/3WUxa6w

Learn about the latest NeMo Retriever NIMs and how they enable developers to deploy models with high accuracy. bit.ly/3SOgIUc

Learn how to achieve the most accurate responses for text Q&A in this upcoming webinar on August 7. bit.ly/4cfKdW1

👀 Members of the hashtag#NVIDIA Developer Program now have free access to downloadable NIM microservices for research, development and testing on up to two GPU nodes. 🎉 Get started: bit.ly/4fxojAw

Check out how the world’s first #generativeAI models for #OpenUSD development will be available as NVIDIA NIM microservices. bit.ly/3WysO56

🦙 The new Llama 3.1 models with the new NVIDIA NeMo Retriever NIMs create a powerful #generativeai duo. ➡️ Technical deep dive on how to build an agentic RAG pipeline by the newest NeMo Retriever embedding and reraking NIM microservices. bit.ly/3WixsDX

🦙 Learn how to build custom #generativeAI models with Llama 3.1 #LLMs through our NVIDIA AI Foundry in our technical deep dive. bit.ly/46jacuj

Join @NealVaidya and I as we dive into production #generativeAI and discover strategies for ensuring data security, compliance, and innovation while managing AI #inference at scale. Register for this upcoming webinar on June 18th to learn more. bit.ly/3Ku5Bv4

Speed time to market and improve TCO by leveraging seamless deployment with NVIDIA NIM, starting with Meta’s Llama 3 70B & 8B, on your preferred cloud service provider, directly accessible from Hugging Face. ➡️ nvda.ws/3VquZZ9. bit.ly/4aMDUsh