BREAKING : Perplexity is working on a new Model Council multi-model system, combining outputs from GPT-5.2, Opus 4.5 and Gemini 3 Pro into one response. In addition, a new mode named Gema is in the works, labelled as "ASI"
LLMS
-
OpenAI Leadership Focus: Capability Over PR Disputes
By
–
Somebody moved his cheese. I use both openAI and Anthropic and reserve the right to make my own choices. But would love it if Sam focuses more on delivering capability than on crying because someone made a kick ass ad.
-
vLLM Kernel Optimizations Boost GB200 Inference Performance
By
–
Impressive deep dive! It’s great to see the vLLM team maximizing the GB200’s potential. These kinds of kernel-level optimizations are exactly why the PyTorch ecosystem continues to be the foundation for next-gen inference performance.
-
Perplexity Launches Advanced Deep Research with SOTA Performance
By
–
Perplexity released a new Advanced Deep Research upgrade with SOTA performance. It scores 79% at Google DeepMind Deep Research QA and tops the newly introduced DRACO Benchmark. Currently available to Max users and expected to arrive for Pro users soon.
-
From March 2025 to Feb 2026: AI writes all code
By
–
Devs, march 2025: Amodei is insane, no way AI writes all our code next year
Devs, feb 2026: -
Mistral Launches Voxtral Transcribe 2 with Affordable API Pricing
By
–
Mini Transcribe 2 via API at $0.003/min.
Realtime via API at $0.006/min, plus open weights. Blogpost: https://
mistral.ai/news/voxtral-t
ranscribe-2
… HuggingFace (weights + realtime demo): https://
huggingface.co/mistralai/Voxt
ral-Mini-4B-Realtime-2602
… Get started: http://
console.mistral.ai/build/audio/sp
eech-to-text
… -
Voxtral Realtime: Ultra-Fast Open Source Streaming Voice Model
By
–
Voxtral Realtime is built for voice agents and live applications. Its natively streaming architecture delivers latency configurable to sub-200ms. And at 480ms, it stays within 1-2% WER of our offline model. We release the model as open weights under Apache 2.0.
-
Grammarly: An AI-Powered Writing Assistant for Improved Grammar and Spelling
By
–
2. Grammarly Un assistant d’écriture propulsé par l’IA qui t’aide à améliorer : la grammaire
l’orthographe
la ponctuation
le vocabulaire Idéal pour écrire mieux, plus vite et sans fautes. -
Mistral AI Releases Voxtral Transcribe 2 Speech-to-Text Models
By
–
Mistral AI released Voxtral Transcribe 2 speech-to-text models and a new Audio Playground on La Platforme.
— 🚨 AI News | TestingCatalog (@testingcatalog) 4 février 2026
Testing time! 👀 https://t.co/nb1KAYmbEH pic.twitter.com/yBq5IpcZLPMistral AI released Voxtral Transcribe 2 speech-to-text models and a new Audio Playground on La Platforme. Testing time!
-
AI Agents Transform PDFs Into Real-Time Intelligence Engine
By
–
What if your documents became a real-time intelligence engine? From the same model family behind our recent ViDoRe V3 win, AI agents built on NVIDIA Nemotron are transforming PDFs and contracts into live insights. See how @EdisonSci
, @Docusign
, and @JusttFintech are