event

Qwen 3.8 Release

13 items · started 2026-09-17 · closed 2026-09-27

  1. Hey everyone, Jovan from UkisAI here! Today, we are introducing Swift, a family of efficient reasoning LLMs based on Qwen, trained by penalizing tokens related to pathological overthinking patterns and restoring accuracy via RL and OPD.

  2. jev-serve Typed probabilistic decisions from any LLM. No text generation.

  3. We introduce Qwen3.8-Omni-Flash, a natively multimodal agentic model for real-world multimodal productivity. Compared with previous omni models, which primarily emphasized perception and interaction, Qwen3.8-Omni-Flash substantially improv…

  4. Hi guys, We recently shipped Astorias AI with he goal of making inference affordable and providing tools for people to build reliable agents. For now we just put qwen 3.8 27b at what we believe to one of the most affordable pricing at $0.1…

  5. Self-Directed-Agent Fully Autonomous agent capable of running 24hrs plus, building complete complex applications (e.g. Postgres-SpringBoot-React) unattended using an RTX5090 and Qwen 3.8 26B Genesis Anchor: 0cdfe09fa3dca3e289184b32aaed265a…

  6. Experiment in agentic autonomy protocols. ~ everything in this repo was created by Qwen 3.8 27B (Q4) running autonomously inside Deepseek Harness, on a single RTX 3090 GPU, for 3 weeks.

  7. Keeping vLLM's Prefix Cache Warm Between Agent Turns From 55% to 95% cached 🔗 I’ve been playing with a few different ways to host Qwen3.8 locally. I’m aiming for something that can replace Claude Code for most of my tasks.

  8. ShapeLearn-Lite Held Up. ShapeLearn Did Better: Qwen 3.8 27B We were a little impatient.

  9. Qwen Studio offers comprehensive functionality spanning chatbot, image and video understanding, image generation, document processing, web search integration, tool utilization, and artifacts.

  10. What are your own experiences on this? What mac/specs do you have?

  11. Hi everybody, we post-trained Qwen 3.8 27B to be more efficient by figuring out which tokens were linked to overthinking and penalizing them without "attacking" the reasoning length directly then fixed the accuracy with a bit of secret sau…

  12. Operating System powered by Qwen 3.8 27B at 1950 tokens/sec! here is what 1,950 tokens/second @Alibaba_Qwen's 3.8 27b actually looks like on @cerebras: i wrote a minimal python web server that turns cerebras inference into a live operatin…

  13. AIC Evals Verified benchmark artifacts, model patches, and integrity receipts for AIC, a closed-source, host-enforced AI software engineering runtime. This repository publishes inspectable evidence from selected AIC evaluation runs.

← all threads