Granite 4.2 3B
Files and memory at 8K context
| Quant | File | Files | + KV + runtime | Smallest fit | Source |
|---|---|---|---|---|---|
| Q4_K_M ibm-granite/granite-4.2-3b-GGUF | 2.09 GiB | 1 | 3.7 GiB | 8 GB GPU | HF files |
| Q5_K_M ibm-granite/granite-4.2-3b-GGUF | 2.43 GiB | 1 | 4.1 GiB | 8 GB GPU | HF files |
| Q8_0 ibm-granite/granite-4.2-3b-GGUF | 3.63 GiB | 1 | 5.3 GiB | 8 GB GPU | HF files |
| FP16 mradermacher/granite-4.2-3b-GGUF | 6.82 GiB | 1 | 8.4 GiB | 12 GB GPU | HF files |
| MLX 4-bit lmstudio-community/granite-4.2-3b-MLX-4bit | 1.92 GiB | 1 | 3.5 GiB | 8 GB Mac | HF files |
| MLX 8-bit lmstudio-community/granite-4.2-3b-MLX-8bit | 3.62 GiB | 1 | 5.2 GiB | 8 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
9.1
Variant: Granite 4.2 3B · observed 2026-10-17 · source
Rank #64 of 75 researched 2026 models.
Arena text rating
1,293
95% CI 1281–1304 · 3,251 votes · published 2026-10-15 · dataset
Exact listed model name; hosted reasoning/quantization may differ from local checkpoint.
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
- Clarification on enabling and evaluating 512K context Hugging Face · 2026-09-10
Qualitative signals only; titles are not verified claims.