Motif 3
Files and memory at 8K context
| Quant | File | Files | + KV + runtime | Smallest fit | Source |
|---|---|---|---|---|---|
| Q4_K_M AesSedai/Motif-3-GGUF | 183 GiB | 5 | unknown | 192 GB GPU (cache unknown) | HF files |
| Q5_K_M AesSedai/Motif-3-GGUF | 220 GiB | 6 | unknown | over 192 GB GPU | HF files |
| MLX 8-bit mixed precision avlp12/Motif-3-Alis-MLX-8bit | 312 GiB | 78 | unknown | 512 GB Mac (cache unknown) | HF files |
No verified file: Q8_0, FP16, BF16, MLX 4-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
33.6 AA estimate
Variant: Motif 3 · observed 2026-10-17 · source
Rank #14 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 5
| Hardware | Quant | Backend | Prompt t/s | Gen t/s | Date | Source |
|---|---|---|---|---|---|---|
| Apple M3 Ultra 512GB | MLX8 affine group64 | avlp12/mlx-lm motif3-support custom fork | — | 37.1 | unknown | source |
| Apple M3 Ultra 512GB | MLX8 affine group64 | avlp12/mlx-lm motif3-support custom fork | — | 24.5 | unknown | source |
| Apple M3 Ultra 512GB | MLX8 affine group64 | avlp12/mlx-lm motif3-support custom fork | — | 14.4 | unknown | source |
| Apple M3 Ultra 512GB | MLX8 affine group64 | avlp12/mlx-lm motif3-support custom fork | 790 | — | unknown | source |
| Apple M3 Ultra 512GB | MLX8 affine group64 | avlp12/mlx-lm motif3-support custom fork | 575 | — | unknown | source |
Reported by the linked sources, not reproduced by LocalFit. All rows in the speed database