K2-Horizon-3.7B
Files and memory at 8K context
| Quant | File | Files | + KV + runtime | Smallest fit | Source |
|---|---|---|---|---|---|
| Q4_K_M IFM/K2-Horizon-3.7B-GGUF | 2.94 GiB | 1 | 5.1 GiB | 8 GB GPU | HF files |
| Q5_K_M IFM/K2-Horizon-3.7B-GGUF | 3.39 GiB | 1 | 5.5 GiB | 8 GB GPU | HF files |
| Q8_0 IFM/K2-Horizon-3.7B-GGUF | 5.02 GiB | 1 | 7.1 GiB | 8 GB GPU | HF files |
| FP16 mradermacher/K2-Horizon-3.7B-GGUF | 9.43 GiB | 1 | 11.6 GiB | 16 GB GPU | HF files |
| MLX 4-bit mixed precision mlx-community/K2-Horizon-3.7B-4bit | 3.85 GiB | 1 | 6.0 GiB | 16 GB Mac | HF files |
| MLX 8-bit mlx-community/K2-Horizon-3.7B-8bit | 5.01 GiB | 2 | 7.1 GiB | 16 GB Mac | HF files |
Sizes are exact file bytes (text weights; projectors listed separately). “+” = recurrent state or other terms unknown. Smallest fit adds 0.5 GiB spare; Macs use 75% of memory.
Quality
Artificial Analysis Intelligence Index
15.6
Variant: K2 Horizon 3.7B · observed 2026-10-17 · source
Rank #41 of 75 researched 2026 models.
Arena text rating
Unknown
Scores describe the evaluated hosted variant, not each quantized file. AA and Arena are never averaged.
Reported speeds 0
No reported measurement yet. Submit one.
Community threads
- Can we have GGUF version Hugging Face · 2026-09-04
- Data mixture Hugging Face · 2026-10-01
Qualitative signals only; titles are not verified claims.