Inkling 2 Small launches on Tinker, benchmarked for post-trainability
simonguozirui · x · 2026-11-06
Inkling 2 Small is live on Thinking Machines' Tinker, with the launch focused on post-trainability: on identical data and algorithms across customer-inspired post-training scenarios, it learns faster than its predecessor, matching Inkling Small's final score on a financial-document summarization task in just 75 training steps. Charts compare finetunability vs. training steps and cost against GLM 5.3 Flash, DeepSeek V4.1 Flash, and Qwen 3.5 397B; reposters urge more OSS model teams to benchmark post-trainability too.
Related event: Thinking Machines Releases Inkling 2 Small with Open Weights Coming(13 posts)→
More from Models
- Thinky Machines launches Inkling 2 Small, scoring 31 on AA Intelligence Index — LiTianleli · 2026-11-06
- Pre-training done in 68 hours with 4-bit used all the way from pre- to post-training — teortaxesTex · 2026-11-06
- METR: No clear bottlenecks seen, capability progress could accelerate further in 12 months — paulnovosad · 2026-11-06
- 18 months from o3 to Claude Mythos: public models still close to last year's frontier — louisvarge · 2026-11-06
- Tencent Hunyuan Hy4 takes #1 Chinese model spot on Agent Arena at $0.57 per task — arena · 2026-11-06
- Agent Arena leaderboard: Tencent's Hy4 leads Chinese models at $0.57/task — arena · 2026-11-06