Thinking Machines ships Inkling 2 Mini: a 26B looped transformer that runs on a laptop
simonguozirui · x · 2026-11-21
Thinking Machines released Inkling 2 Mini, an open-weights looped transformer designed to run locally:
- Its 24 layers include a middle block of 12 layers that loops 3 times per token, letting the model reason internally while emitting fewer tokens
- 26.4B total parameters with 3.4B active; runs on Macs with 32GB RAM; Apache 2.0 license
- Scores 20.1 on the Artificial Analysis Intelligence Index, matching Qwen3.8 27B at roughly 6x lower cost and 4x faster per task
- Hits 48.2% on DeepSWE v1.1, claimed best among open models under 100B parameters
- Because looping can hide reasoning from the chain of thought, the lab tested monitorability: the CoT got shorter but stayed legible with no meaningful drop at this scale
It's a response to feedback that Inkling 2 Small (594B) was too big, positioned as a laptop-runnable model for agentic work.
More from coding & agent
- Redditor wires Claude Code to a spare iPhone for 24/7 real-device agents — MasterReading8368 · 2026-11-21
- Cerebras runs Kimi K3.1 Pro at up to 2,000 tokens/sec, hosts $5,000 hackathon — film_girl · 2026-11-21
- local-ai-registry: one battle-tested local LLM recipe per GPU, with evidence attached — LiTianleli · 2026-11-21
- Sous Labs open-sources Songe, a local fine-tuning engine that trains on your own conversations — alexcovo_eth · 2026-11-21
- Running Claude Code on an in-game GTA phone to approve code on the go — HarshNoiseX · 2026-11-21
- Google Black Friday: Up to 50% Off AI Plans, Antigravity Cache Fix Boosts Quota Work by 20% — op7418 · 2026-11-21