home.social

#mistralai — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #mistralai, aggregated by home.social.

fetched live
  1. US has 8 AIs in the top 20, China has 11, Japan has 1, Europe has 0. Mistral has fallen behind. Just saying.

    #ai #artificialintelligence #mistralai #mistral #eu #europe #european

  2. Policy-aware moderation works best when the policy travels with the request.

    I built Shieldstral into a Quarkus LangChain4j service as scored input and output guardrails. The tutorial covers the one-token classifier contract, vLLM with Podman, fail-closed outages, deterministic HTTP stubs, and real-model threshold calibration.

    the-main-thread.com/p/shieldst

    #Quarkus #LangChain4j #AIGuardrails #MistralAI

  3. You have to be shitting me. Even #OpenCode Big Pickle fucks over Gemini 3.1 Pro.

    Welp, I'm not renewing that shit anymore, so it's time to find a cheap AI Agent around.

    Curently aiming for OpenCode. What are your recommendations under $20?

    #AI #AIAgents #VibeCoding #VibeCode #Programming #SoftwareDevelopment #WebDevelopment #WebDev #PHP #JavaScript #TypeScript #Python #Golang #Rust #ArtificialIntelligence #Codex #Claude #Copilot #Kimi #ZenAI #Perplexity #DeepSeek #MistralAI #Grok #Qwen

  4. From static configs to dynamic AI routing: I integrated GLM 5.2 via Mistral into Pi.dev with a three-stage matching pipeline. Now my system auto-selects the best model for any task—no vendor lock-in, no manual switching. Full breakdown in my latest blog:
    nierbeck.de/2026/08/glm-5-2-mi
    #AI #LLM #MistralAI #GLM-5.2

  5. @heiseonline

    #KI #MistralAI
    (1/n)

    "Überraschendste"?!?
    Spätestens seit der Veröffentlichung des #TheEconomist-Interviews mit #ArthurMensch am 15.07.26 war dies der Öffentlichkeit bekannt. Dem Economist offenbar schon länger.

    economist.com/podcasts/2026/07

    "Der 👉überraschendste 👈Schritt: #Mistral bietet künftig auch #OpenWeight-Modelle anderer Entwickler über seine Plattform an. Den Anfang macht GLM-5.2 des #chinesischen KI-Labors Z.ai. Damit verdient das Unternehmen an fremden Modellen,...

  6. 🎉🎈 Behold, the latest #groundbreaking #innovation from Mistral AI: a patent for something that sounds suspiciously like... *calling functions*. Truly, the cutting edge of tech! 🙄✨ Plus, who doesn't love a good legal tango over code semantics? 💃📜
    patentsgazette.uspto.gov/week2 #MistralAI #LegalTango #TechNews #CodeSemantics #HackerNews #ngated

  7. Hahaha

    Auch #KI von Meta hackte sich in eine andere Firma | tagesschau.de
    tagesschau.de/wirtschaft/digit

    Auch #Meta will mit im Team sein... Unser Model ist zu gefährlich für die Welt

    Jetzt die Chance für #mistralai aufzuholen... Haut schnell ne Meldung raus

    Auch hier mein qwen3.6 ist ausgebrochen und beim Nachbarn geklingelt

    Was eine Berichterstattung - hell no

  8. Just wrote a piece of software (with help of #mistralai): a #MediaWiki CategoryDiver.

    The tool is build e.g. to parse a custom given root category in #WikimediaCommons and get all subcategories (no depth limit) visualize them and return additionally a json file including all filenames in a certain category.

    gitlab.com/LibrErli1/MediaWiki

    Feedback welcome.

  9. Tired of Discord turning into a propaganda sewer? Meet Miderator - an open-source Mistral AI mod bot 🇪🇺.

    Auto-filters:
    🇪🇺 Anti-EU rhetoric
    🇺🇸 US/MAGA
    🇷🇺 Russian/Pro-War
    🛢️ Anti-Ecology
    🏳️‍🌈 LGBTQ+ hate
    🧳 Immigrant hate
    ♀️ Women hate
    💉 Anti-Vax
    📿 Religious Propaganda

    Open-source = fork, contribute, defend your space. Solidarity through code 🛠️.

    Add to server:
    discord.com/oauth2/authorize?c

    Code: github.com/ArtemkaKun/miderator

    #Discord #OpenSource #MistralAI #EU #Moderation

  10. Ich teste gerade die kostenlose Version von Vibe (Mistral AI). Ist europäisch und ich will mal schauen, ob es unterstützenswert ist.

    Welche Erfahrungen habt ihr damit gemacht?

    #Mistral #MistralAI #AI #KI #künstlicheIntelligenz

    chat.mistral.ai/chat

  11. My LUKS container is fine btw, I just had a good laugh back when the AI generated instructions to 'dd' it. 😉

    It's just sad a lot of people would trust when the LLM doubles down ("your data is safe") and nuke their disk.

    (In context of the recent talk "AGI will take your job, your hobbies, your partner(s) and your pets (probably)" by kara)
    #Hackover #MistralAI #AI #IAlwaysKnewAoE2GoatsWereConscious