IFM's open-source 7B model K2-Horizon nearly matches 27B-level performance

IFM released K2-Horizon-7B, a small dense model with its full training pipeline open-sourced, scoring 21 on the AA Intelligence Index and nearly matching Qwen3.6-27B's 22.

2026-09-16 ~ 2026-09-17 · 2 related posts

• IFM's 7B K2-Horizon nearly matches 27B models, tops GPT-5 on BrowseComp, fully open-sourced — karminski3 · 2026-09-16
• IFM's K2-Horizon-7B outscores Qwen3.6-27B despite being a 7B model — rupspace · 2026-09-17