Turn model changes into useful choices: what to try, how to use it, and where it falls short.
Topic editions via RSS ↗Google DeepMind
@GoogleDeepMindSynthID Detector can quickly identify our watermarks in video, audio and images – even after edits, like adding filters. ✂️ You can also use existing AI identification features found in @GoogleChrome, @Google Search, and the @GeminiApp. https://t.co/VglUeteVIL
Google AI
@GoogleAILearn more about identifying AI-generated content: https://t.co/CGOirl0INT
Mistral AI
@MistralAIMeet Mistral Large 4, aka Le Chonk. • 1T parameters, natively multimodal. 49B active. It is the best open weights model from US or Europe on aggregated benchmarks. • State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding. • Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure. • Available to all via API today. Working with cybersecurity partners privately. Open weights release end of October.
Meet Nano Banana 2.1, our latest image generation and editing model 🍌 This upgraded version outperforms our previous models across the board, with notable leaps in visual design, mask-based editing, and subject consistency to help you create more natural-looking images. https://t.co/d858DrzrF3
OpenAI
@OpenAIWe’re releasing a broad range of new mathematical results produced by an internal frontier model. We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results. https://t.co/7N6TPlft1P
Mistral AI
@MistralAIhttps://t.co/S1Kg1eaiuc
Mistral AI
@MistralAIMistral Large 4 is one of the world's strongest AI models for cybersecurity. https://t.co/TzPvFVbO1h
Anthropic
@AnthropicAIWe’re expanding our Cyber Verification Program to give security professionals broader access to our most capable models. Through this program, verified security professionals can access Claude Mythos 5.1, Opus 5.5, and Sonnet 5.5 with safeguards designed for defensive work. We’re also opening up new tiers to allow for authorized offensive work, like penetration testing and red-teaming. https://t.co/blaJtvxhJh
Google DeepMind
@GoogleDeepMindMeet EmbeddingGemma 2, our first natively multimodal open model for on-device embeddings. It expands beyond text to unify code, images, audio, and video in a shared space. 🧵
Mistral AI
@MistralAIRead more here: https://t.co/cvjNTN7YZA
Simon Willison
@simonwMistral can pelican now! https://t.co/Mvtss5e5ay https://t.co/A2GQpvhEF0
Promethean
@PrometheanAIXPROMPT SHARE: She only needs one chance to disappear. A crowded Tokyo station is the perfect place to disappear. Until someone doesn’t let you. One decision changes everything. After that, there’s no time to hesitate. ➜ Video Prompt: Create a 30-second hyper-realistic https://t.co/bwc565KRdz
Simon Willison
@simonwAnyone know when the models started being able to compose competent music? I had Opus 5.5 "write some computer game music for me" and the results were much better than I had expected https://t.co/vFhfgHh7B3
OpenAI
@OpenAIhttps://t.co/edRBkXdUAN
Mistral AI
@MistralAIMistral Large 4 against the field: efficiently reasoning over diverse complex challenges. CTF speedrun: 19 challenges, with tool calls and solve times grounded in actual runs. It solved 18 of the 19. https://t.co/VYOIZDFhE6
Promethean
@PrometheanAIXPROMPT SHARE: Street photography should feel like you interrupted somebody’s night. That’s where tiny unscripted moments suddenly become the entire photograph. ➜ Image Prompt: Photography & Cinematography Create an ultra-realistic nighttime street photograph of a young man https://t.co/XE31KQSRiC
Google DeepMind
@GoogleDeepMindAt 740M parameters, it’s competitive across benchmarks – even outperforming some specialist models more than twice its size. Developers can use it to add multimodal search to their apps – like finding moments in a video using a voice memo – or pair it with Gemma 4 for private, on-device RAG. EmbeddingGemma 2 is released under an Apache 2.0 license. Check out the weights on @HuggingFace and @Kaggle. Find out more → https://t.co/lRsjbvR5VB
Mistral AI
@MistralAIMalware reverse-engineering: solving an out-of-distribution investigation task. When faced with an unknown binary, Mistral Large 4 reverse-engineers it end-to-end. In this case, it concludes the sample is Cobalt Strike, extracts the IoCs and malware configuration, and writes a report with a YARA rule to catch future incidents. A task that could take a day's work, completed in 12 minutes.
Simon Willison
@simonwOne thing I'd LOVE to understand is how Claude "knows" what the music from Secret of Monkey Island sounds like Maybe Anthropic crawled a whole bunch of online music discussion forums with text-based renditions of classic game music?
Simon Willison
@simonwHere's a benchmark that uses Lilypond https://t.co/6yWiT7tG9e
