Frontier AI in your hands. Get work done with @MistralVibe at https://xitter.twitterapi.xyz/t.co/JsGnCVMUFq.
Paris, France
Joined June 2023
- Tweets243
- Following2
- Followers229K
- Likes40
Mistral AI retweeted
"Mistral Large 4"
paper published exclusively on alphaXiv
Their latest technical report shares that aggressively reducing KV-cache memory can hurt an LLM's agentic capabilities, even when its language modeling performance looks unchanged.
Mistral initially reduced KV-cache memory by 16x, but they found that this hurts long-context agentic tasks, where models need to retrieve information from much earlier context.
To fix this, Mistral restored full-context attention to more layers, then adopted DeepSeek Sparse Attention to select the 2,048 most relevant tokens across the context instead of attending to everything.
Their resulting 1T-parameter MoE model (w 52B active) scores 41% on Terminal-Bench 4.0 and 19.7% on FrontierCode 2.0, and its NVFP4-quantized checkpoint fits on four GB300 GPUs.
alphaxiv.org/abs/2611.mistra…
Mistral AI retweeted
Hi @AgentikHQ, Filippo from @MistralAI here.
The model picked STO-3G autonomously, reasoning that it would be fast and that "orbital shapes will be qualitatively right" even if energies are not.
It validated the energies against PySCF across geometries, finding 1e-13 accuracy.
Mistral AI retweeted
Mistral researcher @Sagar_Vaze argues some of the most serious recent AI cyberattacks came from closed models, not open-source ones:
"If you look at the most significant malicious attacks that have happened recently, these have actually been perpetrated not by open weights models, but by closed weights models which have either been jailbroken or, in some cases, during training, escaped their sandbox. In practice, that’s how many of the attacks have been perpetrated."
"Defenders trying to shore up their own defenses are having their requests refused by these same providers, because cutting-edge defensive requests look quite similar to offensive ones."
"People trying to defend from malicious attacks, often from closed weights models, find they have no recourse but to use an open weights model."
"It's critical that there are open weights models of sufficient strength that people can scalably defend their infrastructure."
@MistralAI
Today, we release the weights of a new and improved Le Chonk, advancing the open frontier for agentic coding and cybersecurity.
💻 Built to build: Le Chonk ranks #1 on @cognition’s FrontierCode among open models, and sets a new benchmark for Pareto efficiency on @artificialanalysis.
🛡 Built to defend: State-of-the-art in cyberdefense, put through the paces by the world's leading cyber vendors, including @CrowdStrike, @cloudflare, @rubrik, @zscaler, and others.
Built for sovereignty: Make it yours. Download the weights on @huggingface , customize it with Mistral Forge, and deploy it wherever you need it.
Read more from our team and customers 👇:
Mistral AI retweeted
Props to @MistralAI for the open-weights release of LE CHONK!
Use it in our new Kilo Desktop app today: kilo.ai/install
Today, we release the weights of a new and improved Le Chonk, advancing the open frontier for agentic coding and cybersecurity.
💻 Built to build: Le Chonk ranks #1 on @cognition’s FrontierCode among open models, and sets a new benchmark for Pareto efficiency on @artificialanalysis.
🛡 Built to defend: State-of-the-art in cyberdefense, put through the paces by the world's leading cyber vendors, including @CrowdStrike, @cloudflare, @rubrik, @zscaler, and others.
Built for sovereignty: Make it yours. Download the weights on @huggingface , customize it with Mistral Forge, and deploy it wherever you need it.
Read more from our team and customers 👇:
Mistral AI retweeted
We sat down with @MistralAI's Pierre Stock and Sagar Vaze from Mistral on launch day of Mistral Large 4.
Open weights. Trained in Mistral’s own data centers. Excels at cyber.
Full conversation on cyber, training, evals and sandboxing:
youtu.be/FczbSauVvjw
Mistral AI retweeted
Can an STT model make sense of a 135 year old recording?
I tested @MistralAI new Voxtral transcription model on a noisy French phonograph recording of Gustave Eiffel from 1891.
It matched 100 of 107 words in INA’s published excerpts (6.5% WER), while transcribing 75 additional words omitted from the reference.
It handled Eiffel’s dictation well but struggled with proper names and some guest dialogue.
A fascinating test of modern speech recognition on audio from 1891.
Mistral AI retweeted
🇫🇷 Le Chonk joins the sudo L7 leaderboard 🇫🇷
We just added @MistralAI's Mistral Large 4 to sudo L7, our benchmark for staff-level software engineering.
Mistral Large 4 scores 13.3%, ahead of Muse Spark 1.3 (12.2%) and Grok 4.7 (10.6%).
sudo L7 tests whether coding agents can go beyond writing correct code to exercise the judgment of experienced engineers: navigating messy systems, spotting hidden risks, and knowing when technically correct code is still the wrong thing to ship.
Allez, le Chonk ! 🥖
Leaderboard ↓
surgehq.ai/benchmarks/sudo-l…
Mistral AI retweeted
For their Mistral Large 4 release, @MistralAI ran a blind Surge human eval, with our expert software engineers evaluating five frontier models on coding quality.
Le Chonk ranked #1 among open-weight models and #2 overall, behind only Opus 5.
Unit tests ask: does it work?
Human code review asks: would you merge it?
Our evals measures the difference. Working code is the floor; professional taste and judgment are what make it shippable.
Congrats to the Mistral team on le Chonk and their big release 🇫🇷
Meet 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.
Mistral AI retweeted
Build once. Reuse everywhere.
The Connectors API is now in Public Preview. Register MCP connectors once and use them across Le Chat, AI Studio, and all programmatic tool calls.
Mistral AI retweeted
Since launching Voxtral Realtime, the community response has been remarkable. Today, we share the technical report, launch the Realtime playground in Mistral Studio, and share the model in Hugging Face Transformers. 🧵
Mistral AI retweeted
Mistral Vibe 2.0 is now available on Le Chat Pro and Team plans.
Build, maintain, and ship code faster with the terminal-native coding agent by @MistralAI. Here’s what’s new 🧵
Mistral AI retweeted
One week of @MistralAI's Devstral 2 in Cline.
6.52% diff-edit failure rate.
On the diff-edit chart, that puts it ahead of GLM-4.6 (7.58%) and Kimi-K2 (9.29%) -- while being 123B params (5x smaller than DeepSeek V3.2).
Devstral is FREE during launch so try it out while the promotion lasts.
Mistral AI retweeted
Unmute turns a text LLM into a voice AI. At unmute.sh, it’s @MistralAI's Mistral-Small-3.2-24B, making it fully open-source. Play a quiz game with a snarky host, catch up on tech news, or just hang out and talk. Or modify it to do anything you want!
Mistral AI retweeted
Today we are releasing ether0, our first scientific reasoning model.
We trained Mistral 24B with RL on several molecular design tasks in chemistry. Remarkably, we found that LLMs can learn some scientific tasks more much data-efficiently than specialized models trained from scratch on the same data, and can greatly outperform frontier models and humans on those tasks. For at least a subset of scientific classification, regression, and generation problems, post-training LLMs may provide a much more data-efficient approach than traditional machine learning approaches. 1/n