Source: Document SourceSeptember 3, 2026

K2 Horizon: Six Fully Open Models from 0.9B to 375B Parameters

IFM and MBZUAI released K2 Horizon on September 3 — six fully open Apache 2.0 models ranging from 0.9B to 375B parameters, with weights, training code, data, and checkpoints published.

Key Points:

• All models are released under Apache 2.0 license with full transparency: weights, training code, data, and checkpoints.

• Supported from day one by vLLM, SGLang, Ollama, and Unsloth.

• The smallest tiers claim state-of-the-art results at their respective sizes.

• The range from 0.9B to 375B parameters provides options for different deployment constraints and use cases.

The release provides a fully transparent alternative to proprietary frontier models, with complete reproducibility for research and deployment.

Why It Matters: The fully open release with training data and code provides researchers and enterprises with complete transparency and reproducibility — a counterweight to proprietary model opacity.

K2 Horizon: Six Fully Open Models from 0.9B to 375B Parameters | AI Onboarded