K2-Horizon open-weight family (3.7B-375B, 512K ctx, Apache-2.0) gets day-0 vLLM support
HongyiWang10 · x · 2026-09-03
vLLM announced day-0 support for K2-Horizon: IFM released not just final weights but intermediate checkpoints, detailed data-construction recipes, training code, and fine-grained logs. The family spans six models from 0.9B/3.7B dense to a 375B-A23B MoE, all Apache-2.0 with 512K context and reasoning/tool-calling baked into checkpoints for agents and long documents — prototype small, scale up without changing stacks. The post also notes vLLM-Omni now serves MiniMax H3, an open-weight joint video/audio diffusion transformer served via an OpenAI-compatible /v1/videos API, generating H.264 MP4 with native stereo audio and supporting t2va, first/last-frame, and omni-reference editing modes.
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