Thinking Machines Launches Inkling Small, Beating Full-Size Model in Benchmarks
soumitrashukla9 · x · 2026-08-01
Thinking Machines released the Inkling Small model. With 276B total parameters and 12B active parameters trained on Nvidia GB300 NVL72 systems, it is roughly a quarter the size of the full Inkling model.
Despite its smaller footprint, it outperforms the full-size model on several benchmarks: scoring 46.6% vs. 46.0% on Humanity's Last Exam with tools, and 83.4% vs. 79.8% on IFBench. The company attributes this to improved pre-training data and recipes specifically tailored for the smaller model.
However, it clearly lags in factual recall, scoring 20.9% on Simple QA Verified compared to 43.9% for the full Inkling. Additionally, it features native reasoning over audio and images, variable thinking effort, and a 1 million token context window.
Related event: Inkling-Small: 276B params, 12B active, open-source tops ARC(48 posts)→
More from Models
- User spots model openly generating "ideal responses" to train other models — liminal_bardo · 2026-10-16
- Reflection AI open-sources Beam model weights under Apache 2.0 with 40+ partners — victormustar · 2026-10-16
- Reflection's 501B MoE Beam hits 463 tok/s on 8x GB200 with NVFP4 at day 0 — zhyncs42 · 2026-10-16
- JevBench v1.7.46 leaderboard: new d1-ultra tops 32 API systems at 76.1 composite — airesearch12 · 2026-10-16
- LMArena launches Agent Arena and Code Arena leaderboards for model comparison — arena · 2026-10-16
- Third-party tests find NVIDIA Nemotron-3.5 Bolt underperforms nano on pandas/numpy and decision-making tasks — norenEnmotalen · 2026-10-16