home.social

#mistralvibe — Public Fediverse posts

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

fetched live
  1. Debugging with Mistral: Learning, Not Just Fixing

    Depending on which social bubble one is into, people are drawn to letting LLMs do everything. Latest tools promise to fully configure or even manage entire server setups. While that sounds amazing (and I want to try it out some time) that approach has some downsides: Loss of control: I don’t really know what’s happening under the hood. If something breaks, I’m left in the dark. Loss of Learning: I want to know / learn how and why things (don't) work. Digital Sovereignty: Relying on an external connection is creating a hard dependency that I want to avoid. And hosting my own LLM isn't an option as my mini PC doesn’t have the RAM or power for that. So, while these tools are impressive, they’re not for me — at least not now or not in my private useacse of maintaining my main home-server. But, that's no reason to not leverage it at all! […]

    locked.de/debugging-with-mistr

  2. Rate my setup: Altes #Samsung bzw. #android Tablet > #termux > #ssh my-dev-server > #Docker > #mistralvibe #KI

    Wenn das nur noch Gedanken lesen könnte, könnte ich wortwörtlich im Schlaf coden.
    Bis dahin muss ich mich mit Spracheingabe zufriedengeben. 😎

    #ratemysetup

  3. It's a shame that there are so many people who don't know #mistralvibe ...

    #mistral

  4. so apparently you can use a Mistral API key in a coding agent, such as Pi, and have it use your monthly subscription but it is tricky to find.

    There is a separate dedicated Vibe API key which is separate from the pay-as-you-go keys.

    On the Code Extensions page, click on advanced, and look for "Vibe API Key"

    chat.mistral.ai/code/extensions

    #mistral #mistralai #mistralvibe #pidev

  5. @simsa0.bsky.social

    The answer above about the grading was just a joke, of course. This is how I do it: anyone who has used #MistralVibe gets a better grade. After all, we're Europeans - and this is the final showdown.

  6. Mistral AI Surges with Enterprise Focus and European Ambitions

    Mistral AI, a French company, has released Mistral Vibe, an AI platform for businesses. It helps automate tasks and offers data control for European companies.

    #MistralAI, #AIEurope, #BusinessAI, #MistralVibe, #TechNews

    newsletter.tf/mistral-ai-launc

  7. Mistral AI, a European AI firm, has launched a new platform called Mistral Vibe to help businesses automate tasks. This move aims to compete with major AI companies by offering more data control.

    #MistralAI, #AIEurope, #BusinessAI, #MistralVibe, #TechNews
    newsletter.tf/mistral-ai-launc

  8. I haven't been using JetBrains Webstorm for a while. Now they've finally made it possible to use Mistral Vibe in Jetbrains.
    Which warms my battered electronic EU heart. 🤖 🇪🇺 ❤️

    #mistral #mistralvibe #JetBrains

  9. Echt schade, dass #mistral und #mistralvibe so unpräsent sind... Ich bin doch immer wieder überrascht

  10. GenAI coding needs more than just a Licence

    I just found this blog post from Rob Bowley in my RSS feed and one paragraph just so resonated with what I read and hear so much:

    For CEOs and founders hoping to benefit, the answer isn’t as simple as handing out Claude licences […]. It’s investing in the engineering culture and practices. Unglamorous, slow work, but there’s no way around it.

    Rob Bowley

    I’ve been trying #MistralVibe […]

    locked.de/genai-coding-needs-m #Claude #GithubCopilot #Mistral #MistralVibe #SoftwareEngineering
  11. It can hardly become more apparent that code assistants aren't for free when you see the receipt popup on your phone while sending a prompt.

    Had that yesterday when I was using #mistralvibe when my account balance crossed the minimum value so that they could book.

    #codeassistant

  12. Evening-shootout:
    A user reported that an invite-mail that I sent shows raw HTML instead of the rendered HTML.

    Fired up #githubCopilot and #mistralVibe in plan mode and let search.

    Both found the bug and proposed the same (right) solution.
    Just #Mistral was quicker.

  13. I'm using #GithubCopilot and #MistralVibe to create a small personal Web application.

    Today I needed a *forgot password* function.
    So I let GHC plan, checked the plan carefully, adjusted the plan, and finally let it write code and tests.
    Over 200 tests passed.
    Let Mistral do a code review, all fine.

    First manual test in the browser: in one form, the token wasn't handed over...

    My learnings:
    - I likely wouldn't have noticed this bit in a manual review, too
    - but two agents didn't notice it (that's very sad)
    - I'm thinking about adding selenium tests - for a small personal project... Am I overdoing this? Maybe. But it could be pretty cool, though!! 😀

  14. Irgendwie sind #mistral Agents etwas widerborstiger als #githubcopilot Agents 🤔
    Oder ich mache etwas falsch..

    #mistralvibe

  15. Why Settle for One AI Assistant When You Can Have Two?

    Two weeks ago, I discovered that Mistral.ai also provides a coding assitant, similar to GitHub Copilot (GHC), called Mistral Vibe (GitHub page).

    In those two weeks I’ve been using Mistral Vibe in parallel to GHC. Just because I wanted to try and see the difference! And just after a couple of days I noticed that the agents definition in Mistral Vibe are a bit different from GHC (in hindsight: […]

    locked.de/why-settle-for-one-a #Agentic #AI #GithubCopilot #MistralVibe #SoftwareEngineering
  16. RE: norden.social/@maik/1163289481

    Artikel wie schon von Arstechnica. Die interessantere Frage ist wie lange es dauern wird bist auf dieser Basis Angriffe auf / mit / durch Claude Code gefahren werden. Und was für Effekte daraus zu sehen sein werden.

    Abgesehen davon: Ich bin hier um mit Ihnen über das Wort Go...äh, OpenCode zu reden: opencode.ai/

    Gleiches LLM coding, allerdings was Modelle angeht offen und inhärent gleich FOSS als Codebasis. Funktioniert auch mit Mistral, ist deutlich weiter entwickelt als Mistrals VibeCLI.
    Kritik und Vorteile alle wie gehabt, mehr Auswahl und mehr Offenheit.

    #ai #claude #claudecode #opencode #mistral #mistralvibe #codestral #devstral2 #devstral #llm #vibecoding

  17. I started experimenting with #opencode and #mistralvibe for some part of my work, and either I'm dumb or missing something, but can we somehow see individual changes inside #neovim?

    I'm pretty sure I saw some solution in past, but can't find it now, and when I recently opened undotree after changes made by Vibe, it was empty and even my previous changes weren't in undotree

    #mistral #vibecoding #ai #gemini #geminicli #claudecode (last two because solution I saw was about them)

  18. FUCK

    I just had my first "LLMs can actually do shit" moment when it comes to software development. I've created two Python scripts since before that took the output from the meshtastic cli application and rendered some nice output.

    Today I just asked - still the local small LLM that is in no way as capable as the large models - for the LLM to refactor those two Python scripts into a full fledged backend service with a web frontend and a few bells and whistles too.

    ... and it just did it. I had to fix some extremely minor stuff but all in all, yes, this was mostly just a background thing running on my own local hardware for a few hours today.

    Well ok. I have a nefarious reason for doing this. Once complete I will then attack the security of the application/service. I expect there to be issues, but my confidence has been slightly shaken just now.

    #MistralVibe

  19. If you're interesting in using Mistral Vibe for LLM-aided development - here's a tidbit of info that will make your life easier:

    The context length you set in Vibe's config needs to be some X amount of tokens _lower_ than the largest context your hosted LLM supports.

    Vibe will run automatic compact on the history when you're close to maxing out the context length, but it will not manage to stay below the setting so if you don't do this your hosted LLM will run out of memory.

    #MistralVibe

  20. 🚀 #Devstral2 & #MistralVibe CLI: State-of-the-art #opensource agentic coding models #AI #coding #developers

    🔥 #Devstral2 (123B) achieves 72.2% on SWE-bench Verified — SOTA for open-weight code agents under modified #MIT license

    #DevstralSmall2 (24B) scores 68.0% on SWE-bench, runs locally on consumer hardware under #Apache2 license

    💰 Up to 7x more cost-efficient than Claude Sonnet at real-world tasks

    🧵 👇

  21. #Mistral AI launched #Devstral2, a new #codingmodel, and #MistralVibe, a command-line interface for #codeautomation. Devstral 2, available in two sizes, emphasises context awareness and production-grade workflows. Mistral is valued at €11.7 billion and partnered with Kilo Code and Cline to release Devstral 2. techcrunch.com/2025/12/09/mist #tech #media #news