K2 Horizon 3.7B
Description
K2 Horizon 3.7B is the small dense member of IFM's (MBZUAI) fully open K2 Horizon family: a 3.7B-core decoder-only reasoning model (36 layers; about 5.1B parameters in the released safetensors including its 250K-token vocabulary embeddings) pretrained on 22.9T tokens, with a native 524,288-token (512K) context window from midtraining onward. Text in, text out, with per-request reasoning effort (high recommended) and JSON/XML tool-call formats. Released 2026-09-03 under Apache 2.0; on 2026-10-01 the main branch was updated to an anti-doom-loop (FTPO) checkpoint, and the model-card scores were refreshed on 2026-10-28 (they differ from the launch-blog figures). BF16 weights plus a GGUF build, intermediate checkpoints, training data, code, and logs. No priced provider is cataloged yet.
Capability Radar
Science is estimated from LLM Stats science scores or reasoning when no dedicated science benchmarks are available.
Rankings
| Domain | #Rank | Score | Source |
|---|---|---|---|
| Agentic Capability | 72 | 47.0 | LS |
Benchmark Scores (LLM Stats)
(LLM Stats (zeroeval))Agents
Math
Reasoning
Tool Calling
AA Evaluation Indices
(Artificial Analysis)No AA evaluation data available
LLM Stats Category Scores
(LLM Stats (zeroeval))Pricing
No pricing data available
Speed
No speed data available
Provider Price Ranking
No provider data available