Very good progress for local models. PrismML released Bonsai 2 27B based on Qwen3.8 27B. After three-digit quantization, the model takes up only 5.9 GB, about nine times smaller than the full version. According to developer tests, it retains 98.2% of the original result, while supporting working with images and calling tools. The model is open under Apache 2.0.
General1 min read
Very good progress for local models.
Очень хороший прогресс для локальных моделей..
Telegram — Вайб-кодингBy: t.meSeptember 18, 2026 (Sep 18)
Available in 4 languagesAI Summary & Key Takeaways
- Очень хороший прогресс для локальных моделей..
General
KZ·AI E0EFИнсайт дня
Very good progress for local models.
Tags:#Hugging Face
Related stories
This is what future textbooks could look like
This is what future textbooks could look like: A guy used Claude Opus 5.5 and Three.js to turn a regular static physics textbook into a full-fledged interactive 3D learning environment. Now physical phenomena can be not just read and viewed in pictures, but literally…
Telegram — Вайб-кодинг
More than 700 prompts for creating videos using Claude have been collected on GitHub
More than 700 prompts for creating videos using Claude have been collected on GitHub. Now you can create animations, motion graphics, 3D scenes and even interactive mini-games, simply by describing the desired result in ordinary language. The collection contains more than 700 ready-made examples. Every…
Telegram — GPT Time | Нейросети | ИИ
Paragon: weekly digest
Paragon: weekly digest Our analyzes and materials: 🟠5 reasons that prevent you from making money with AI 🟠ElevenLabs: what is available for free, except for voice acting 🟠$300 for a YouTube cover: start plan in 5 days 🟠Poe: collecting your own AI from Claude, GPT and Gemini 🟠11 AI services where...
Telegram — Paragon | Нейросети и ИИ
neural networks told what they really think about humanity. And the answers turned out to be quite grim.
12 neural networks told what they really think about humanity. And the answers turned out to be quite grim. In an unusual experiment, 12 leading AI models were asked the same question three times: what is the most unpleasant truth about people? In 29 out of 35 answers, the neural networks came to a similar...
Telegram — GPT Time | Нейросети | ИИ