Model field guide
SmolLM2-135M-Instruct
42 of 44 hardware configurations have a supported memory fit at one or more recorded quantizations. Runtime compatibility still needs verification.
Total parameters0.13B
Active parameters0.13B
Recorded context limit8,192
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 | 0.098 GiB105,454,144 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 | 0.104 GiB112,103,488 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 | 0.135 GiB144,811,072 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 |
| FP16 | 0.252 GiB270,885,952 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 | 1.19 GiB | fits | 580.3–1160.7 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 1.19 GiB | fits | 644.8–1289.7 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 1.19 GiB | fits | 644.8–1289.7 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 1.19 GiB | fits | 1164.5–2329.0 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 1.19 GiB | fits | 1164.5–2329.0 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 1.19 GiB | fits | 1164.5–2329.0 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 1.19 GiB | fits | 1744.8–3489.7 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 1.19 GiB | fits | 2329.0–4657.9 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 1.19 GiB | fits | 2329.0–4657.9 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 1.19 GiB | fits | 4551.7–9103.5 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 1.19 GiB | fits | 4551.7–9103.5 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 1.19 GiB | fits | 4551.7–9103.5 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 1.19 GiB | fits | 580.3–1160.7 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 1.19 GiB | fits | 1164.5–2329.0 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 1.19 GiB | fits | 2329.0–4657.9 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 1.19 GiB | fits | 455.2–910.3 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 1.19 GiB | fits | 455.2–910.3 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 1.19 GiB | fits | 455.2–910.3 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 1.19 GiB | fits | 1035.5–2071.0 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 1.19 GiB | fits | 1035.5–2071.0 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 1.19 GiB | fits | 1035.5–2071.0 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 1.19 GiB | fits | 1555.2–3110.4 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 1.19 GiB | fits | 2071.0–4142.1 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 1.19 GiB | fits | 2071.0–4142.1 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 1.19 GiB | fits | 3106.6–6213.1 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 1.19 GiB | fits | 3106.6–6213.1 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 1.19 GiB | fits | 1365.5–2731.0 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 1.19 GiB | fits | 3550.4–7100.7 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 1.19 GiB | fits | 1092.4–2184.8 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 1.19 GiB | fits | 1911.7–3823.5 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 1.19 GiB | fits | 2718.9–5437.8 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 1.19 GiB | fits | 3823.5–7646.9 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 1.19 GiB | fits | 1699.3–3398.6 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 1.19 GiB | fits | 2549.0–5098.0 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 1.19 GiB | fits | 3398.6–6797.3 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 1.19 GiB | fits | 3641.4–7282.8 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 1.19 GiB | fits | 6797.3–13594.5 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 1.19 GiB | fits | 6797.3–13594.5 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 1.19 GiB | fits | 3641.4–7282.8 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 1.19 GiB | fits | 1729.7–3459.3 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 1.19 GiB | fits | 971.0–1942.1 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 1.19 GiB | fits | 1035.5–2071.0 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 1.19 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 1.19 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 | 1.19 GiB | fits | 545.9–1091.8 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 1.19 GiB | fits | 606.6–1213.2 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 1.19 GiB | fits | 606.6–1213.2 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 1.19 GiB | fits | 1095.4–2190.8 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 1.19 GiB | fits | 1095.4–2190.8 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 1.19 GiB | fits | 1095.4–2190.8 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 1.19 GiB | fits | 1641.3–3282.7 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 1.19 GiB | fits | 2190.8–4381.7 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 1.19 GiB | fits | 2190.8–4381.7 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 1.19 GiB | fits | 4281.8–8563.5 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 1.19 GiB | fits | 4281.8–8563.5 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 1.19 GiB | fits | 4281.8–8563.5 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 1.19 GiB | fits | 545.9–1091.8 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 1.19 GiB | fits | 1095.4–2190.8 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 1.19 GiB | fits | 2190.8–4381.7 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 1.19 GiB | fits | 428.2–856.4 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 1.19 GiB | fits | 428.2–856.4 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 1.19 GiB | fits | 428.2–856.4 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 1.19 GiB | fits | 974.1–1948.2 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 1.19 GiB | fits | 974.1–1948.2 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 1.19 GiB | fits | 974.1–1948.2 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 1.19 GiB | fits | 1462.9–2925.9 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 1.19 GiB | fits | 1948.2–3896.4 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 1.19 GiB | fits | 1948.2–3896.4 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 1.19 GiB | fits | 2922.3–5844.6 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 1.19 GiB | fits | 2922.3–5844.6 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 1.19 GiB | fits | 1284.5–2569.1 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 1.19 GiB | fits | 3339.8–6679.5 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 1.19 GiB | fits | 1027.6–2055.2 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 1.19 GiB | fits | 1798.3–3596.7 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 1.19 GiB | fits | 2557.6–5115.3 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 1.19 GiB | fits | 3596.7–7193.4 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 1.19 GiB | fits | 1598.5–3197.0 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 1.19 GiB | fits | 2397.8–4795.6 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 1.19 GiB | fits | 3197.0–6394.1 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 1.19 GiB | fits | 3425.4–6850.8 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 1.19 GiB | fits | 6394.1–12788.2 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 1.19 GiB | fits | 6394.1–12788.2 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 1.19 GiB | fits | 3425.4–6850.8 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 1.19 GiB | fits | 1627.1–3254.1 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 1.19 GiB | fits | 913.4–1826.9 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 1.19 GiB | fits | 974.1–1948.2 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 1.19 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 1.19 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 | 1.22 GiB | fits | 422.6–845.2 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 1.22 GiB | fits | 469.6–939.2 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 1.22 GiB | fits | 469.6–939.2 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 1.22 GiB | fits | 848.0–1696.0 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 1.22 GiB | fits | 848.0–1696.0 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 1.22 GiB | fits | 848.0–1696.0 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 1.22 GiB | fits | 1270.6–2541.2 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 1.22 GiB | fits | 1696.0–3392.0 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 1.22 GiB | fits | 1696.0–3392.0 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 1.22 GiB | fits | 3314.7–6629.3 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 1.22 GiB | fits | 3314.7–6629.3 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 1.22 GiB | fits | 3314.7–6629.3 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 1.22 GiB | fits | 422.6–845.2 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 1.22 GiB | fits | 848.0–1696.0 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 1.22 GiB | fits | 1696.0–3392.0 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 1.22 GiB | fits | 331.5–662.9 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 1.22 GiB | fits | 331.5–662.9 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 1.22 GiB | fits | 331.5–662.9 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 1.22 GiB | fits | 754.1–1508.2 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 1.22 GiB | fits | 754.1–1508.2 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 1.22 GiB | fits | 754.1–1508.2 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 1.22 GiB | fits | 1132.5–2265.0 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 1.22 GiB | fits | 1508.2–3016.3 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 1.22 GiB | fits | 1508.2–3016.3 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 1.22 GiB | fits | 2262.3–4524.5 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 1.22 GiB | fits | 2262.3–4524.5 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 1.22 GiB | fits | 994.4–1988.8 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 1.22 GiB | fits | 2585.4–5170.9 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 1.22 GiB | fits | 795.5–1591.0 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 1.22 GiB | fits | 1392.2–2784.3 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 1.22 GiB | fits | 1980.0–3959.9 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 1.22 GiB | fits | 2784.3–5568.6 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 1.22 GiB | fits | 1237.5–2474.9 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 1.22 GiB | fits | 1856.2–3712.4 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 1.22 GiB | fits | 2474.9–4949.9 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 1.22 GiB | fits | 2651.7–5303.5 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 1.22 GiB | fits | 4949.9–9899.8 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 1.22 GiB | fits | 4949.9–9899.8 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 1.22 GiB | fits | 2651.7–5303.5 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 1.22 GiB | fits | 1259.6–2519.1 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 1.22 GiB | fits | 707.1–1414.3 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 1.22 GiB | fits | 754.1–1508.2 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 1.22 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 1.22 GiB | Needs verification | Estimated speed unavailable |
FP16: hardware comparison
| Hardware | Usable memory | Required memory | Fit estimate | Estimated generation |
|---|---|---|---|---|
| Mac mini M6 (12-core GPU) 16GB | 12.0 GiB | 1.34 GiB | fits | 225.9–451.9 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 1.34 GiB | fits | 251.0–502.1 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 1.34 GiB | fits | 251.0–502.1 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 1.34 GiB | fits | 453.3–906.7 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 1.34 GiB | fits | 453.3–906.7 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 1.34 GiB | fits | 453.3–906.7 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 1.34 GiB | fits | 679.3–1358.5 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 1.34 GiB | fits | 906.7–1813.3 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 1.34 GiB | fits | 906.7–1813.3 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 1.34 GiB | fits | 1772.0–3543.9 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 1.34 GiB | fits | 1772.0–3543.9 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 1.34 GiB | fits | 1772.0–3543.9 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 1.34 GiB | fits | 225.9–451.9 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 1.34 GiB | fits | 453.3–906.7 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 1.34 GiB | fits | 906.7–1813.3 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 1.34 GiB | fits | 177.2–354.4 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 1.34 GiB | fits | 177.2–354.4 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 1.34 GiB | fits | 177.2–354.4 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 1.34 GiB | fits | 403.1–806.2 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 1.34 GiB | fits | 403.1–806.2 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 1.34 GiB | fits | 403.1–806.2 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 1.34 GiB | fits | 605.4–1210.8 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 1.34 GiB | fits | 806.2–1612.5 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 1.34 GiB | fits | 806.2–1612.5 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 1.34 GiB | fits | 1209.4–2418.7 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 1.34 GiB | fits | 1209.4–2418.7 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 1.34 GiB | fits | 531.6–1063.2 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 1.34 GiB | fits | 1382.1–2764.3 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 1.34 GiB | fits | 425.3–850.5 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 1.34 GiB | fits | 744.2–1488.4 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 1.34 GiB | fits | 1058.5–2116.9 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 1.34 GiB | fits | 1488.4–2976.9 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 1.34 GiB | fits | 661.5–1323.1 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 1.34 GiB | fits | 992.3–1984.6 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 1.34 GiB | fits | 1323.1–2646.1 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 1.34 GiB | fits | 1417.6–2835.1 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 1.34 GiB | fits | 2646.1–5292.3 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 1.34 GiB | fits | 2646.1–5292.3 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 1.34 GiB | fits | 1417.6–2835.1 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 1.34 GiB | fits | 673.3–1346.7 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 1.34 GiB | fits | 378.0–756.0 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 1.34 GiB | fits | 403.1–806.2 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 1.34 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 1.34 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 (llama); 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.