#open-weight — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #open-weight, aggregated by home.social.
-
💥 MiniMax M3 enters the global top 10 (#8, 54.7 pts) with 4x fewer parameters than competitors — just weeks after M2.7 (#20).
Kimi K2.6 (256K ctx) and DeepSeek V4 (1M ctx): free and competitive.
Continuing to pay — and share your data — for OpenAI/Antropic is becoming a real strategic question.
-
💥 MiniMax M3 enters the global top 10 (#8, 54.7 pts) with 4x fewer parameters than competitors — just weeks after M2.7 (#20).
Kimi K2.6 (256K ctx) and DeepSeek V4 (1M ctx): free and competitive.
Continuing to pay — and share your data — for OpenAI/Anthropic is becoming a real strategic question.
-
@BijanBowen on YT!
Bijan Bowen
Ornith 1.0 First Look & Test – The BEST New Local Coding Models?
-
@BijanBowen on YT!
Bijan Bowen
Ornith 1.0 First Look & Test – The BEST New Local Coding Models?
-
Why the spread of powerful local AI models makes copyright enforcement even harderThe central theme of Walled Culture the book (free digital versions available) is the growing incompatibility between traditional copyright law and the digital, online world. The culmination of that process was the EU Copyright Directive, passed in 2019. The Directive was supposed to be transposed into local laws by 2021, but a year after that deadline, many EU member states had failed to do […]
#ai #article15 #cars #china #eu #euCopyrightDirective #fragmentation #genai #generativeAi #localModels #marketplace #openSource #openWeight #overblocking #redFlagAct #transparency #uk #uploadFilters https://walledculture.org/why-the-spread-of-powerful-local-ai-models-makes-copyright-enforcement-even-harder/ -
Following an order by the US government, #Anthropic has suspended access to #Fable5 and #Mythos5 by all foreign nationals. In case we need another reason to switch to #openweight models: nobody can "suspend access" to #Deepseek, #Kimi or #Qwen, once they run in hundreds of EU servers.
-
The LLM with the lowest hallucination rate is an #openweight model: #MinimaxM3, which was released a few days ago. Hey almighty #Opus48, where are you?!
-
🚀 MiniMax M3 porta 1 milione di token in open weight: contesti lunghissimi, accesso più aperto e modelli chiusi sotto pressione. #AI #OpenWeight
-
Für alle, die sich aufgrund der neuesten @freakshow Ausgabe einmal Open Weight LLMs ausprobieren, aber nicht direkt viel Geld dafür ausgeben möchten, kann ich #OpenCode Go sehr empfehlen. Es ist erstaunlich zuverlässig und schnell für 10$/Monat. Hier gibt es aktuell noch 5 $ Guthaben: https://opencode.ai/go?ref=DQ3GSMX4PW
-
Für alle, die sich aufgrund der neuesten @freakshow Ausgabe einmal Open Weight LLMs ausprobieren, aber nicht direkt viel Geld dafür ausgeben möchten, kann ich #OpenCode Go sehr empfehlen. Es ist erstaunlich zuverlässig und schnell für 10$/Monat. Hier gibt es aktuell noch 5 $ Guthaben: https://opencode.ai/go?ref=DQ3GSMX4PW
-
Ho chiesto a una #AI un programma dettagliato per una settimana in una città europea.
Il modello #openweight Kimi K2.6 mi ha risposto dopo 5 minuti da un server europeo, usando 20mila token. Ha consumato 53 mL di acqua💧 e prodotto 10 g di CO2. 💨
So queste cose perché ho usato #MeliousAI, un servizio di una giovane startup tedesca che sto provando in questi giorni.
Usare gli LLM in modo consapevole passa anche dalla trasparenza!
-
Ho chiesto a una #AI un programma dettagliato per una settimana in una città europea.
Il modello #openweight Kimi K2.6 mi ha risposto dopo 5 minuti da un server europeo, usando 20mila token. Ha consumato 53 mL di acqua💧 e prodotto 10 g di CO2. 💨
So queste cose perché ho usato #MeliousAI, un servizio di una giovane startup tedesca che sto provando in questi giorni.
Usare gli LLM in modo consapevole passa anche dalla trasparenza!
-
I have just discovered #MeliousAI and I am impressed. These guys are building exactly what I think #Europe should build: not an European #LLM, but infrastructures and services that allow to run open models in a secure way.
This is what Melious is doing: they run #openweight models on EU servers, offering API, integrations with existing coding tools, and a chat interface where chat history is stored locally on the browser. No data ever leaves Europe. Plus, they provide estimates of CO2 emission and energy/water consumption for every prompt.
Open models are cheaper, and they now perform as well as the more famous proprietary counterparts. In my opinion, this is the right path for #SovereignAI in Europe. Well done, MeliousAI! 💪🏻
-
Moving away from expensive frontier models ( #OpenAI, #Claude, #Gemini) to build a custom #openweight AI setup. My current workflow orchestrates #kimi k2.6, #deepSeek v4, and #glm using Oh My OpenAgent as base.
Read about my setup here: https://www.richardorilla.website/seting_up_opencode.html
-
Mistral Medium 3.5 Developer Guide: API, Remote Agents & Pricing 2026
Mistral Medium 3.5 is a 128B open-weight dense model with 77.6% SWE-Bench Verified and a 256K context window, released April 29, 2026 under a modified MIT license. It ships with...
https://wowhow.cloud/blogs/mistral-medium-3-5-developer-guide-api-remote-agents-2026
-
Today I coded for 4 hours with #OpenCode using GLM 5 instead of Claude Sonnet 4.6. The project is a #Deno app written in #typescript and the agent had to refactor code and HTML layout.
I perceived no difference in capabilities. GLM5 handled edits well, made some mistakes just like Sonnet, and produced output of the same quality.
But GLM 5 is 3x smaller than Sonnet!
Private AI models are wasteful. They're too big, use too much power, and are too expensive to run. #openweight is the future.