Post-Cutoff.com
  1. Home
  2. Timeline
  3. 2026
  4. Tim Dettmers' dlab open-source week: 1.5-bit inference, a…

Tim Dettmers' dlab open-source week: 1.5-bit inference, a 125B model on one 24 GB GPU, CliffCompaction and a local research agent

★★after cutoffopen-sourcedlabCarnegie Mellon Universityconfidence: medium

On Sept 21, 2026 quantization researcher Tim Dettmers announced "dlab Open Source Week: Frontier AI on Your Own Hardware": two open-source projects and four papers released over the following days. Headline claims: a 125B Qwen 3.8 Flash Next running on a single 24 GB GPU, a Qwen 3.6 35B-A3B at 1.5 bits per weight reaching ~450 tokens/s on Apple Metal, and a private beta of bitsandbytes2. There is also CliffCompaction, an auto-compaction proxy for long coding-agent runs (arXiv 2609.26779), and a local autonomous-research agent he says beats frontier labs' deep-research systems.

Key facts

What happened

Tim Dettmers, author of QLoRA and the bitsandbytes library, used a week of releases from his lab to argue that near-frontier AI can run on hardware people own. The pieces were an inference framework (with a Mac/Metal path), aggressive low-bit compression through bitsandbytes2, an agent harness, CliffCompaction for very long agent sessions, and an autonomous research agent. The post drew attention on Hacker News (185 points).

Why it matters

If the claims hold, 1.5-bit weights and single-consumer-GPU inference of 100B+ MoE models push open-weights AI further out of datacenters. That matters both for access and for the debate about controlling open models. The comparison with frontier deep-research systems comes from the author and should be treated as unverified.

Changelog

  • 2026-09-30: created (from leads queue; only the CliffCompaction paper was checked on arXiv, the other three papers were not located)

Related posts (1)

Sources (5)

id: 2026-09-21-dettmers-dlab-open-source-week · updated 2026-09-30 · open in the interactive timeline