When was K2 Horizon 375B-A23B released?

K2 Horizon 375B-A23B was released on September 3, 2026 by ifm. This is the official K2 Horizon 375B-A23B release date tracked on LLM Stats.

How big is K2 Horizon 375B-A23B?

K2 Horizon 375B-A23B has 375 billion parameters. It was trained on 15.1 trillion tokens. It ships as an open-weight model, so you can download and run it on your own hardware.

Who created K2 Horizon 375B-A23B?

K2 Horizon 375B-A23B was created by ifm.

What is the license for K2 Horizon 375B-A23B?

K2 Horizon 375B-A23B is released under the Apache 2.0 license. This is an open-source / open-weight license that permits self-hosting.

Where is the K2 Horizon 375B-A23B paper or technical report?

K2 Horizon 375B-A23B has a paper or technical report available at https://ifm.ai/blog/k2/. Use that source for architecture, training, release and evaluation details.

What models should I compare K2 Horizon 375B-A23B against?

Common K2 Horizon 375B-A23B comparisons include K2 Horizon 375B-A23B vs GPT-5 High, K2 Horizon 375B-A23B vs MiMo-V2.5-Pro, K2 Horizon 375B-A23B vs Step-3.5-Flash. Compare them side by side for benchmark scores, pricing, context window, latency and provider availability.

K2 Horizon 375B-A23B Benchmarks, Pricing & Context Window

Name: K2 Horizon 375B-A23B
Author: ifm

K2 Horizon 375B-A23B: Benchmarks, Pricing & Context Window

K2 Horizon 375B-A23B is a language model from Institute of Foundation Models, released in September 2026.

K2 Horizon 375B-A23B is the flagship of the Institute of Foundation Models' (MBZUAI) fully open K2 Horizon family: a sparse Mixture-of-Experts reasoning model with 375B total and 23B active parameters per token (61 layers, 192 routed

K2 Horizon 375B-A23B benchmarks

Capability tiers

Standing within each category, adjusted for leaderboard depth.

Real tasks performance

High-confidence performance for K2 Horizon 375B-A23B across real-world prompt categories. Only 95% intervals at most 4 points wide are shown.

Performance by conversation depth

How K2 Horizon 375B-A23B holds up as conversations get longer.

Quality Tracker

K2 Horizon 375B-A23B Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

llm-stats.com - Fri Oct 30 2026

Notice missing or incorrect data?

K2 Horizon 375B-A23B model size

K2 Horizon 375B-A23B has 375 billion parameters and was trained on 15.1 trillion tokens. See how it compares to other models in the same parameter range.

ParametersTraining tokens

375BMoE

15.1Ttokens

40× tokens-to-params ratio

Frontier (200B+)

375B

1B7B70B405B

Try now

huggle

K2 Horizon 375B-A23Bin Huggle

Make it with
K2 Horizon 375B-A23B.

K2 Horizon 375B-A23B latency

K2 Horizon 375B-A23B time to first token, sustained output throughput, and failed-request rate from live model usage over the trailing 7 days.

K2 Horizon 375B-A23B examples

Recent arena outputs from K2 Horizon 375B-A23B, picked from the highest-ranked matchups.

K2 Horizon 375B-A23B license

K2 Horizon 375B-A23B is released under the Apache 2.0 license, which permits commercial use, has 375.0B parameters.

License: Apache 2.0; Commercial use allowed
Parameters: 375.0B

Apache License 2.0 - allows commercial use

K2 Horizon 375B-A23B resources

Official sources for K2 Horizon 375B-A23B: official launch post, source repository, model weights.

K2 Horizon 375B-A23B vs other models

The most-compared alternatives to K2 Horizon 375B-A23B are GPT-5 High, MiMo-V2.5-Pro, Step-3.5-Flash. Open any pair side-by-side for benchmarks, pricing, context, and latency.