Model field guide
granite-4.2-3b
42 of 44 hardware configurations have a supported memory fit at one or more recorded quantizations. Runtime compatibility still needs verification.
Total parameters3.66B
Active parameters3.66B
Recorded context limit131,072
Architecturedense
Verified 2026-10-22 · license metadata: apache-2.0 · commercial use: yes (review the terms).
Actual files, not a size shortcut.
| Quantization | Total file size | Source |
|---|---|---|
| Q4_K_M | 2.090 GiB2,244,011,552 bytes; 1 file(s)Sum of exact file byte sizes in the repository listing at the recorded commit; every declared shard is present. Auxiliary vision projector and speculative-decoding files are excluded. Labels follow publisher filenames; mixed-precision expert tensors may remain in a differently labeled file, so the estimator uses measured bytes rather than assumed bits per weight. | Pinned file recordVerified 2026-10-22 |
| Q5_K_M | 2.434 GiB2,613,634,592 bytes; 1 file(s)Sum of exact file byte sizes in the repository listing at the recorded commit; every declared shard is present. Auxiliary vision projector and speculative-decoding files are excluded. Labels follow publisher filenames; mixed-precision expert tensors may remain in a differently labeled file, so the estimator uses measured bytes rather than assumed bits per weight. | Pinned file recordVerified 2026-10-22 |
| Q8_0 | 3.625 GiB3,892,651,552 bytes; 1 file(s)Sum of exact file byte sizes in the repository listing at the recorded commit; every declared shard is present. Auxiliary vision projector and speculative-decoding files are excluded. Labels follow publisher filenames; mixed-precision expert tensors may remain in a differently labeled file, so the estimator uses measured bytes rather than assumed bits per weight. | Pinned file recordVerified 2026-10-22 |
The hardware picture.
Q4_K_M: hardware comparison
| Hardware | Usable memory | Required memory | Fit estimate | Estimated generation |
|---|---|---|---|---|
| Mac mini M6 (12-core GPU) 16GB | 12.0 GiB | 3.40 GiB | fits | 27.3–54.5 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 3.40 GiB | fits | 30.3–60.6 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 3.40 GiB | fits | 30.3–60.6 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 3.40 GiB | fits | 54.7–109.4 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 3.40 GiB | fits | 54.7–109.4 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 3.40 GiB | fits | 54.7–109.4 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 3.40 GiB | fits | 82.0–164.0 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 3.40 GiB | fits | 109.4–218.9 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 3.40 GiB | fits | 109.4–218.9 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 3.40 GiB | fits | 213.9–427.8 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 3.40 GiB | fits | 213.9–427.8 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 3.40 GiB | fits | 213.9–427.8 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 3.40 GiB | fits | 27.3–54.5 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 3.40 GiB | fits | 54.7–109.4 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 3.40 GiB | fits | 109.4–218.9 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 3.40 GiB | fits | 21.4–42.8 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 3.40 GiB | fits | 21.4–42.8 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 3.40 GiB | fits | 21.4–42.8 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 3.40 GiB | fits | 48.7–97.3 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 3.40 GiB | fits | 48.7–97.3 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 3.40 GiB | fits | 48.7–97.3 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 3.40 GiB | fits | 73.1–146.2 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 3.40 GiB | fits | 97.3–194.7 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 3.40 GiB | fits | 97.3–194.7 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 3.40 GiB | fits | 146.0–292.0 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 3.40 GiB | fits | 146.0–292.0 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 3.40 GiB | fits | 64.2–128.3 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 3.40 GiB | fits | 166.8–333.7 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 3.40 GiB | fits | 51.3–102.7 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 3.40 GiB | fits | 89.8–179.7 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 3.40 GiB | fits | 127.8–255.5 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 3.40 GiB | fits | 179.7–359.4 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 3.40 GiB | fits | 79.9–159.7 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 3.40 GiB | fits | 119.8–239.6 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 3.40 GiB | fits | 159.7–319.4 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 3.40 GiB | fits | 171.1–342.2 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 3.40 GiB | fits | 319.4–638.9 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 3.40 GiB | fits | 319.4–638.9 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 3.40 GiB | fits | 171.1–342.2 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 3.40 GiB | fits | 81.3–162.6 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 3.40 GiB | fits | 45.6–91.3 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 3.40 GiB | fits | 48.7–97.3 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 3.40 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 3.40 GiB | Needs verification | Estimated speed unavailable |
Q5_K_M: hardware comparison
| Hardware | Usable memory | Required memory | Fit estimate | Estimated generation |
|---|---|---|---|---|
| Mac mini M6 (12-core GPU) 16GB | 12.0 GiB | 3.75 GiB | fits | 23.4–46.8 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 3.75 GiB | fits | 26.0–52.0 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 3.75 GiB | fits | 26.0–52.0 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 3.75 GiB | fits | 47.0–94.0 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 3.75 GiB | fits | 47.0–94.0 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 3.75 GiB | fits | 47.0–94.0 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 3.75 GiB | fits | 70.4–140.8 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 3.75 GiB | fits | 94.0–187.9 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 3.75 GiB | fits | 94.0–187.9 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 3.75 GiB | fits | 183.7–367.3 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 3.75 GiB | fits | 183.7–367.3 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 3.75 GiB | fits | 183.7–367.3 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 3.75 GiB | fits | 23.4–46.8 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 3.75 GiB | fits | 47.0–94.0 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 3.75 GiB | fits | 94.0–187.9 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 3.75 GiB | fits | 18.4–36.7 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 3.75 GiB | fits | 18.4–36.7 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 3.75 GiB | fits | 18.4–36.7 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 3.75 GiB | fits | 41.8–83.6 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 3.75 GiB | fits | 41.8–83.6 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 3.75 GiB | fits | 41.8–83.6 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 3.75 GiB | fits | 62.7–125.5 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 3.75 GiB | fits | 83.6–167.1 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 3.75 GiB | fits | 83.6–167.1 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 3.75 GiB | fits | 125.3–250.7 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 3.75 GiB | fits | 125.3–250.7 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 3.75 GiB | fits | 55.1–110.2 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 3.75 GiB | fits | 143.2–286.5 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 3.75 GiB | fits | 44.1–88.2 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 3.75 GiB | fits | 77.1–154.3 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 3.75 GiB | fits | 109.7–219.4 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 3.75 GiB | fits | 154.3–308.5 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 3.75 GiB | fits | 68.6–137.1 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 3.75 GiB | fits | 102.8–205.7 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 3.75 GiB | fits | 137.1–274.3 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 3.75 GiB | fits | 146.9–293.8 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 3.75 GiB | fits | 274.3–548.5 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 3.75 GiB | fits | 274.3–548.5 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 3.75 GiB | fits | 146.9–293.8 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 3.75 GiB | fits | 69.8–139.6 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 3.75 GiB | fits | 39.2–78.4 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 3.75 GiB | fits | 41.8–83.6 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 3.75 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 3.75 GiB | Needs verification | Estimated speed unavailable |
Q8_0: hardware comparison
| Hardware | Usable memory | Required memory | Fit estimate | Estimated generation |
|---|---|---|---|---|
| Mac mini M6 (12-core GPU) 16GB | 12.0 GiB | 4.94 GiB | fits | 15.7–31.4 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 4.94 GiB | fits | 17.5–34.9 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 4.94 GiB | fits | 17.5–34.9 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 4.94 GiB | fits | 31.5–63.1 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 4.94 GiB | fits | 31.5–63.1 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 4.94 GiB | fits | 31.5–63.1 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 4.94 GiB | fits | 47.3–94.5 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 4.94 GiB | fits | 63.1–126.2 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 4.94 GiB | fits | 63.1–126.2 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 4.94 GiB | fits | 123.3–246.6 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 4.94 GiB | fits | 123.3–246.6 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 4.94 GiB | fits | 123.3–246.6 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 4.94 GiB | fits | 15.7–31.4 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 4.94 GiB | fits | 31.5–63.1 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 4.94 GiB | fits | 63.1–126.2 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 4.94 GiB | fits | 12.3–24.7 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 4.94 GiB | fits | 12.3–24.7 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 4.94 GiB | fits | 12.3–24.7 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 4.94 GiB | fits | 28.1–56.1 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 4.94 GiB | fits | 28.1–56.1 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 4.94 GiB | fits | 28.1–56.1 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 4.94 GiB | fits | 42.1–84.3 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 4.94 GiB | fits | 56.1–112.2 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 4.94 GiB | fits | 56.1–112.2 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 4.94 GiB | fits | 84.2–168.3 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 4.94 GiB | fits | 84.2–168.3 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 4.94 GiB | fits | 37.0–74.0 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 4.94 GiB | fits | 96.2–192.4 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 4.94 GiB | fits | 29.6–59.2 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 4.94 GiB | fits | 51.8–103.6 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 4.94 GiB | fits | 73.7–147.3 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 4.94 GiB | fits | 103.6–207.2 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 4.94 GiB | fits | 46.0–92.1 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 4.94 GiB | fits | 69.1–138.1 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 4.94 GiB | fits | 92.1–184.1 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 4.94 GiB | fits | 98.6–197.3 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 4.94 GiB | fits | 184.1–368.3 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 4.94 GiB | fits | 184.1–368.3 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 4.94 GiB | fits | 98.6–197.3 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 4.94 GiB | fits | 46.9–93.7 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 4.94 GiB | fits | 26.3–52.6 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 4.94 GiB | fits | 28.1–56.1 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 4.94 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 4.94 GiB | Needs verification | Estimated speed unavailable |
A “fits” result checks estimated memory only. “Needs verification” preserves missing cache, vision, or placement information. All speed figures are estimated from the published bandwidth formula.
Field-level sources and verification dates
- id: Official source · verified 2026-10-22. Lowercase URL slug derived from the upstream repository ID; baseModel retains the exact Hugging Face identifier.
- name: Official source · verified 2026-10-22
- baseModel: Official source · verified 2026-10-22
- downloads: Official source · verified 2026-10-22
- tasks: Official source · verified 2026-10-22. Chat denotes a text-generating model, including base models that may require a chat template or instruction tuning. Code label is inferred from the repository model-family name. Vision uses pipeline metadata or vision_config.
- license: Official source · verified 2026-10-22
- commercialUse: Official source · verified 2026-10-22. Official upstream organization metadata identifies a permissive standard license. Commercial use is permitted subject to the license terms, notices, and applicable restrictions; this is a metadata-based classification, not an individualized legal approval.
- license_url: Official source · verified 2026-10-22. No recognized standalone license file; linked pinned model card.
- architecture: Official source · verified 2026-10-22. Classification from model_type and expert/state-space configuration fields (granite); hybrid classification takes precedence over expert count.
- parametersB: Official source · verified 2026-10-22. Exact safetensors parameter total reported by the Hugging Face API, divided by 1e9.
- activeParametersB: Official source · verified 2026-10-22. Dense model: all parameters are included. This is not a runtime FLOP measurement.
- contextLength: Official source · verified 2026-10-22. Configured maximum token positions; not a guarantee of runtime support or quality at that length. No speculative RoPE scaling applied.
- kvCache: Official source · verified 2026-10-22. Standard attention KV shape from num_hidden_layers, num_key_value_heads, and head_dim or hidden_size / num_attention_heads; FP16 cache bytes are a separate estimator choice.
- licenseStandard: Official source · verified 2026-10-22. Official standard license text supporting the permissive-license classification; model applicability is identified separately by upstream metadata.