IFM releases K2 Horizon: six fully open models from 0.9B to 375B with 512K context
rupspace · x · 2026-09-03
IFM AI launched K2 Horizon, a family of six open-weight models from 0.9B to 375B, billed as the largest fully open-source model release ever — code, training data, and recipes all public. The flagship 375B targets reasoning and agentic work with a native 512K context, matching open MoE models up to 2.6x its size on tool use, terminal, and long-horizon tasks. Smaller models set SOTA at their scales (0.9B runs on a watch, 7B on a phone), and SGLang offers Day-0 support.
Related event: MBZUAI's IFM Releases K2 Horizon: Six Fully Open Models from 0.9B to 375B(30 posts)→
More from Models
- Users want open-format export from Claude Docs as charts only save as images — BlackHC · 2026-10-20
- EMNLP study: LLM safety interventions drop up to 4.5x when distress is wrapped in delusion — MaartenSap · 2026-10-20
- Cybersecurity-focused VLM OrcaCyber-Nano-1.0 trends on Hugging Face — orcarouter · 2026-10-20
- 0.8B local model with 2,000 labels beats Jev's frontier API on banking data — kalyan_kpl · 2026-10-20
- New TTS benchmark tests 26 models across 30,564 requests: fastest at 46ms — mattturck · 2026-10-20
- Agents handle unreleased language Jai flawlessly, even compile-time execution — lucasmeijer · 2026-10-20