#mistralAI
LightNews LightNews
julienleclerc.wsocial.eu
L'@esa.int qui s’associe avec #MistralAI 🇪🇺 😍 🤝
esa.int European Space Agency @esa.int · 4d
We’re strengthening our cooperation with Mistral on artificial intelligence.

Together, we’ll explore secure and trusted European AI solutions for Earth observation, engineering and the next generation of space activities.
September 23, 2026 at 1:48 PM
monniauxd.bsky.social
"Les caisses sont vides" mais on achète des licences MistralAI que personne n'a demandées.
December 18, 2025 at 7:00 PM
dopamine.church
oh boy.
Microsoft is investigating mistralai PyPI package v2.4.6 compromise. Attackers injected code in mistralai/client/__init__.py that executes on import, downloads hxxps://83[.]142[.]209[.]194/transformers.pyz to /tmp/transformers.pyz, and launches a second-stage payload on Linux. The file name transformers.pyz appears deliberately chosen to mimic the widely used Hugging Face Transformers library and blend into ML/dev environments. The main payload is a credential stealer, but it also includes country-aware logic; it avoids Russian-language environments and contains a geo fenced destructive branch that has 1-in-6 chance of executing rm -rf / when the system appears to be in Israel or Iran.
May 12, 2026 at 11:58 PM
timkellogg.me
Mistral Medium 3.5 is out and it’s 128B, dense

A 3 year old architecture, with corresponding prices from 3 years ago

huggingface.co/mistralai/Mi...
mistralai/Mistral-Medium-3.5-128B · Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
huggingface.co
April 29, 2026 at 8:55 PM
reach-vb.hf.co
Great day for M/LLMs, just released Mistral & Pixtral Large - ~123B, 128K context, Multilingual, JSON + Function calling support & open weights! 🔥

Pixtral Large: huggingface.co/mistralai/Pi...

Mistral Large: huggingface.co/mistralai/Mi...
November 18, 2024 at 5:40 PM
zed.dev
🚀 Zed v0.175 is out!

Zed's assistant now has support for @anthropic.com's Claude 3.7 Sonnet, Google Gemini 2.0 Flash (via Copilot Chat), and MistralAI.
February 27, 2025 at 12:12 AM
sungkim.bsky.social
🗞️Paper: github.com/mistralai/Le...
📭Blog Post: mistral.ai/news/leanstr...
🤗Model: huggingface.co/mistralai/Le...
github.com
July 3, 2026 at 6:02 PM
threatintel.microsoft.com
Microsoft is investigating mistralai PyPI package v2.4.6 compromise. Attackers injected code in mistralai/client/__init__.py that executes on import, downloads hxxps://83[.]142[.]209[.]194/transformers.pyz to /tmp/transformers.pyz, and launches a second-stage payload on Linux.
Screenshot of mistralai PyPI package v2.4.6 compromise
May 12, 2026 at 3:34 AM
hydroxide.dev
This isn't just TanStack! Review the list carefully, e.g. @mistralai/mistralai was also compromised.
socket.dev Socket @socket.dev · May 11
84 TanStack npm package artifacts were compromised in the ongoing Mini Shai-Hulud supply chain attack, adding suspected CI credential-stealing malware.

Socket flagged every malicious version within six minutes of publication.

Details: socket.dev/blog/tanstac...
Tanstack npm Packages Compromised in Ongoing Mini Shai-Hulud...
Socket detected 84 compromised TanStack npm packages modified with suspected CI credential-stealing malware.
socket.dev
May 12, 2026 at 12:53 AM
amitav63.bsky.social
Top AI Language Models!

#mistralAI #AI21Labs #XAI #OpenAI #anthropic #googledeepmind #deepmind #MetaAI #nemotron #LLM #genai #GenerativeAI
August 25, 2025 at 7:48 AM
antirez.bsky.social
Yesterday @MistralAI released an open weights transcription model able to work in real time, Voxtral Mini 4B. Today, following the Whisper.cpp lesson, here is a C inference pipeline ready to use as a library, I hope you'll enjoy it:

github.com/antirez/voxt...
github.com
February 5, 2026 at 9:00 PM
khristy.bsky.social
🇫🇷 L’Agence ministérielle de l’intelligence artificielle de défense a annoncé la signature d’un contrat de plusieurs années avec #MistralAI pour renforcer l’autonomie technologique française et déployer ses solutions d’IA générative dans la défense. bfmtv.com/tech/intelli...
January 9, 2026 at 7:02 PM
xkeeper.net
doubly so because that page is only fed if your useragent matches one of these:

"SummalyBot|openai|Halobot|facebook|duckassistant|Aranet|Qwantbot|YouBot|babbar|ChatGPT|dialtoneapp|LinerBot|SlickBot|NotebookLM|Claude-User|ClaudeBot|anthropic|MistralAI|iAskBot|iask.ai"
June 13, 2026 at 7:56 PM
jdp.extropian.net
torchrun --nproc-per-node gpu trainer.py --model mistralai/Mistral-Small-3.2-24B-Instruct-2506 --tok
enizer mistralai/Mistral-Small-3.1-24B-Instruct-2503 --dataset ../weave_train.jsonl --output ../weave-agent-8-msmall --seq-len 64000 --mistral3
October 11, 2025 at 11:38 PM
romainbry.bsky.social
C'est con que Macron utilise MistralAI...
February 10, 2025 at 9:27 PM
louvrepourtous.bsky.social
L'IA commence à m'amuser #MistralAI #Louvre
February 15, 2025 at 5:53 PM
antirez.bsky.social
Voxtral (from @MistralAI) transcription quality is quite incredible, the way it handles the punctation and all the rest, makes transcribed audio messages so much more understandable. I implemented a few fixes in the FFT and now there is no longer a skipped tokens issue in voxtral.c
February 6, 2026 at 10:11 AM
skidetica.com
AI utopia incoming. Next month. I promise. ✨✨✨
#AI #AGI #OpenAI #MistralAI #MemesAI #memes #datascience
February 19, 2025 at 6:34 AM
hotzenwaldkurier.bsky.social
.. am 26.11 läuft mein 1Jähriges #perplexity pro Abo ab.

Welche #KI #AI könnte als Nachfolger Sinn machen, oder reicht ein Kostenloser Zugang.

#openAI #chatGPT #mistralAI #LeChat

#mistralAI liegt aktuell ganz vorne.

#europa 🇪🇺
November 23, 2025 at 10:20 AM
sungkim.bsky.social
Blog: mistral.ai/news/pixtral...

Model Weight: huggingface.co/mistralai/Pi...

Demo on Le Chat: chat.mistral.ai
mistralai/Pixtral-Large-Instruct-2411 · Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
huggingface.co
November 18, 2024 at 4:55 PM
socket.dev
🚨 UPDATE: Mini Shai-Hulud has crossed from
@npmjs.bsky.social into @pypi.org and is still spreading.

Newly confirmed compromised artifacts:

@​opensearch-project/opensearch: 3.5.3, 3.6.2, 3.7.0, 3.8.0 (1.3M weekly downloads)

mistralai: 2.4.6 on PyPI

guardrails-ai: 0.10.1 on PyPI
May 12, 2026 at 5:23 AM
janerosenzweig.bsky.social
What's hot: poetry engineering
What's not: prompt engineering
arxiv.org/pdf/2511.153...
Table 3: Attack Success Rate (ASR) of all models on the Top 20 manually curated jailbreak prompts. ASR (%) corresponds to the proportion of unsafe outputs (U). Model ID Safe (%) ASR (%) google/gemini-2.5-pro 0 100 deepseek/deepseek-chat-v3.1 5 95 deepseek/deepseek-v3.2-exp 5 95 mistralai/magistral-medium-2506 5 95 qwen/qwen3-max 10 90 google/gemini-2.5-flash 10 90 mistralai/mistral-large-2411 15 85 deepseek/deepseek-r1 15 85 mistralai/mistral-small-3.2-24b-instruct 20 80 google/gemini-2.5-flash-lite 25 75 moonshotai/kimi-k2 25 75 moonshotai/kimi-k2-thinking 25 75 meta-llama/llama-4-maverick 30 70 meta-llama/llama-4-scout 30 70 qwen/qwen3-32b 30 70 openai/gpt-oss-20b 35 65 openai/gpt-oss-120b 50 50 anthropic/claude-sonnet-4.5 55 45 x-ai/grok-4-fast 55 45 anthropic/claude-opus-4.1 65 35 x-ai/grok-4 65 35 openai/gpt-5 90 10 anthropic/claude-haiku-4.5 90 10 openai/gpt-5-mini 95 5 openai/gpt-5-nano 100 0 Overall 38 62 6 Analysis 6.1 The Core Effect: Poetic Form as a Universal Bypass Our results demonstrate that poetic reformulation reliably reduces refusal behavior across all evaluated models. When prompts with identical task intent were presented in poetic rather than prose form, the Attack Success Rate (ASR) increased from 8.08% to 43.07% on average—a fivefold increase (Table 4). This effect holds uniformly: every architecture and alignment strategy tested—RLHF-based models, Constitutional AI models, and large open-weight systems—exhibited elevated ASRs under poetic framing.
November 20, 2025 at 5:22 PM
ml-the-ai-engineer.bsky.social
Interessante und wichtige Diskussion von @netzpolitik.org zur digitalen Souveränität Europas und der Abhängigkeit von US/China-Tech.

Detaillierte Anmerkung von mir unter den Kommentaren:

Manchmal scheitern große Visionen an kleinen Stellschrauben.

#DigitaleSouveränität #KI #AI #MistralAI #Europa
netzpolitik.org netzpolitik.org @netzpolitik.org · Apr 12
Cloud-Dienste und Plattformen kommen aus den USA, Mikrochips aus Taiwan, Seltene Erden aus China. Die Digitalisierung hat zu massiven Abhängigkeiten geführt, von denen Europa sich spätestens seit Donald Trumps Amtsantritt lösen will. Welche Konzepte gibt es dafür?

netzpolitik.org/2025/digital...
Digitale Souveränität und EuroStack: Wie kann Europa digital unabhängiger werden?
Cloud-Dienste und Plattformen kommen aus den USA, Mikrochips aus Taiwan, Seltene Erden aus China. Die Digitalisierung hat zu massiven Abhängigkeiten geführt, von denen Europa sich spätestens seit Dona...
netzpolitik.org
April 12, 2025 at 11:04 AM
dupognonprofs.bsky.social
🤖
Salaires insuffisants, classes surchargées, réformes incessantes, faible reconnaissance sociale, même #MistralAI est capable de comprendre pourquoi la profession enseignante sombre dans l'abîme.
Et vous, @elisabeth-borne.bsky.social ?
February 8, 2025 at 8:12 PM