Files and memory at 8K context
| Quant | File | Files | + KV + runtime | Smallest fit | Source |
|---|---|---|---|---|---|
| Q4_K_M mradermacher/G9v3-3B-GGUF | 1.72 GiB | 1 | 3.1 GiB | 8 GB GPU | HF files |
| Q5_K_M mradermacher/G9v3-3B-GGUF | 1.99 GiB | 1 | 3.4 GiB | 8 GB GPU | HF files |
| Q8_0 mradermacher/G9v3-3B-GGUF | 2.96 GiB | 1 | 4.4 GiB | 8 GB GPU | HF files |
| FP16 mradermacher/G9v3-3B-GGUF | 5.57 GiB | 1 | 7.0 GiB | 8 GB GPU | HF files |
| MLX 4-bit cof139/G9v3-3B-mlx-4Bit | 1.57 GiB | 1 | 3.0 GiB | 8 GB Mac | HF files |
No verified file: MLX 8-bit. 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
10.8 AA estimate
Variant: G9v3-3B · observed 2026-10-17 · source
Rank #58 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.