Back to News

PrismML releases Bonsai 2 27B, compressing Qwen3.8 27B to 5.9 GB for smartphones

#model compression#prismml#qwen#smartphone ai

PrismML has released Bonsai 2 27B, a compressed version of Alibaba's Qwen3.8 27B model. The model fits in 5.9 GB, making it suitable for smartphones, while retaining 98.2% of Qwen's benchmark scores.

Coverage timeline

  1. Techmeme
    PrismML releases Bonsai 2 27B, which compresses Alibaba's Qwen3.8 27B to 5.9 GB, small enough for smartphones, while retaining 98.2% of Qwen's benchmark scores (Julie Bort/TechCrunch)

    Julie Bort / TechCrunch : PrismML releases Bonsai 2 27B, which compresses Alibaba's Qwen3.8 27B to 5.9 GB, small enough for smartphones, while retaining 98.2% of Qwen's benchmark scores — If AI lab PrismML isn't on your radar yet, it should be — not because it's raised gobs of money (it hasn't yet …

Related stories

  • Alibaba releases Qwen3.8-Flash, an open-weight 125B-param model on Qwen 4 architecture

    Alibaba released Qwen3.8-Flash, an open-weight 125B-parameter multimodal mixture-of-experts model built on its next-gen Qwen 4 architecture, claiming it rivals Claude Opus 4.6 and V4-Flash. The model has only 6B active parameters, and Qwen Office launched a standard mode powered by it, reporting ~100% faster generation and 75% lower token consumption in real office-task tests. The release also serves as an early preview of the Qwen 4 architecture.

    6 reports · 5 sources697 points · 232 comments

  • Qwen3.8 27B quantization benchmark: 4-bit holds, 1-bit collapses

    A Hacker News post benchmarks quantizations of the Qwen3.8 27B model, reporting that 4-bit quantization maintains performance well, while 1-bit quantization leads to a collapse in quality. The post provides practical data for researchers and engineers considering model compression.

    1 reports · 1 sources284 points · 139 comments

  • Alibaba Updates Qwen3.8-Max, Tops CodeArena Front-End Coding Leaderboard

    On September 2, 2026, Alibaba updated its flagship Qwen3.8-Max model, with performance gains after specialized post-training for coding and office tasks. The new version scored 1691 on the CodeArena WebDev leaderboard, a 22-point increase, ranking first ahead of models like Claude Opus5 and Kimi K3. The model, with 2.4 trillion parameters and 1 million token context, is now available via API on the Qwen AI platform and integrated into Qwen Office, Qoder, and Qwen App.

    2 reports · 2 sources

  • Zhipu AI open-sources GLM-5.3-Flash, a 320B multimodal model priced at 1/40 of Claude Opus 4.8

    Zhipu AI released and open-sourced GLM-5.3-Flash, the first natively multimodal model in the GLM-5 series, with 320B total parameters (18B active). It scores 57 on the Artificial Analysis Intelligence Index, matching Claude Opus 4.8, and is priced at one-tenth of GLM-5.3, or one-fortieth of Opus 4.8. The model uses a hybrid sparse and linear attention architecture, runs on domestic Chinese chips, and was tested anonymously as Ox-Alpha on OpenRouter and OpenCode.

    8 reports · 5 sources1126 points · 573 comments

News updates every hour. Subscribe in your RSS reader to get new stories automatically.

Subscribe