Model field guide
Qwen2.5-1.5B-Instruct
42 of 44 hardware configurations have a supported memory fit at one or more recorded quantizations. Runtime compatibility still needs verification.
Total parameters1.54B
Active parameters1.54B
Recorded context limit32,768
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 | 1.041 GiB1,117,320,736 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 | 1.197 GiB1,285,494,304 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 | 1.764 GiB1,894,532,128 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 | 3.316 GiB3,560,416,288 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 | 2.15 GiB | fits | 54.8–109.5 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 2.15 GiB | fits | 60.9–121.7 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 2.15 GiB | fits | 60.9–121.7 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 2.15 GiB | fits | 109.9–219.8 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 2.15 GiB | fits | 109.9–219.8 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 2.15 GiB | fits | 109.9–219.8 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 2.15 GiB | fits | 164.7–329.4 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 2.15 GiB | fits | 219.8–439.6 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 2.15 GiB | fits | 219.8–439.6 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 2.15 GiB | fits | 429.6–859.2 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 2.15 GiB | fits | 429.6–859.2 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 2.15 GiB | fits | 429.6–859.2 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 2.15 GiB | fits | 54.8–109.5 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 2.15 GiB | fits | 109.9–219.8 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 2.15 GiB | fits | 219.8–439.6 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 2.15 GiB | fits | 43.0–85.9 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 2.15 GiB | fits | 43.0–85.9 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 2.15 GiB | fits | 43.0–85.9 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 2.15 GiB | fits | 97.7–195.5 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 2.15 GiB | fits | 97.7–195.5 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 2.15 GiB | fits | 97.7–195.5 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 2.15 GiB | fits | 146.8–293.6 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 2.15 GiB | fits | 195.5–390.9 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 2.15 GiB | fits | 195.5–390.9 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 2.15 GiB | fits | 293.2–586.4 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 2.15 GiB | fits | 293.2–586.4 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 2.15 GiB | fits | 128.9–257.8 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 2.15 GiB | fits | 335.1–670.2 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 2.15 GiB | fits | 103.1–206.2 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 2.15 GiB | fits | 180.4–360.9 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 2.15 GiB | fits | 256.6–513.2 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 2.15 GiB | fits | 360.9–721.7 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 2.15 GiB | fits | 160.4–320.8 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 2.15 GiB | fits | 240.6–481.2 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 2.15 GiB | fits | 320.8–641.5 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 2.15 GiB | fits | 343.7–687.4 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 2.15 GiB | fits | 641.5–1283.1 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 2.15 GiB | fits | 641.5–1283.1 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 2.15 GiB | fits | 343.7–687.4 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 2.15 GiB | fits | 163.2–326.5 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 2.15 GiB | fits | 91.6–183.3 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 2.15 GiB | fits | 97.7–195.5 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 2.15 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 2.15 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 | 2.31 GiB | fits | 47.6–95.2 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 2.31 GiB | fits | 52.9–105.8 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 2.31 GiB | fits | 52.9–105.8 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 2.31 GiB | fits | 95.5–191.1 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 2.31 GiB | fits | 95.5–191.1 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 2.31 GiB | fits | 95.5–191.1 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 2.31 GiB | fits | 143.1–286.3 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 2.31 GiB | fits | 191.1–382.1 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 2.31 GiB | fits | 191.1–382.1 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 2.31 GiB | fits | 373.4–746.8 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 2.31 GiB | fits | 373.4–746.8 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 2.31 GiB | fits | 373.4–746.8 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 2.31 GiB | fits | 47.6–95.2 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 2.31 GiB | fits | 95.5–191.1 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 2.31 GiB | fits | 191.1–382.1 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 2.31 GiB | fits | 37.3–74.7 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 2.31 GiB | fits | 37.3–74.7 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 2.31 GiB | fits | 37.3–74.7 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 2.31 GiB | fits | 84.9–169.9 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 2.31 GiB | fits | 84.9–169.9 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 2.31 GiB | fits | 84.9–169.9 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 2.31 GiB | fits | 127.6–255.2 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 2.31 GiB | fits | 169.9–339.8 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 2.31 GiB | fits | 169.9–339.8 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 2.31 GiB | fits | 254.8–509.7 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 2.31 GiB | fits | 254.8–509.7 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 2.31 GiB | fits | 112.0–224.0 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 2.31 GiB | fits | 291.2–582.5 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 2.31 GiB | fits | 89.6–179.2 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 2.31 GiB | fits | 156.8–313.7 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 2.31 GiB | fits | 223.0–446.1 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 2.31 GiB | fits | 313.7–627.3 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 2.31 GiB | fits | 139.4–278.8 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 2.31 GiB | fits | 209.1–418.2 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 2.31 GiB | fits | 278.8–557.6 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 2.31 GiB | fits | 298.7–597.4 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 2.31 GiB | fits | 557.6–1115.2 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 2.31 GiB | fits | 557.6–1115.2 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 2.31 GiB | fits | 298.7–597.4 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 2.31 GiB | fits | 141.9–283.8 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 2.31 GiB | fits | 79.7–159.3 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 2.31 GiB | fits | 84.9–169.9 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 2.31 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 2.31 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 | 2.87 GiB | fits | 32.3–64.6 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 2.87 GiB | fits | 35.9–71.8 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 2.87 GiB | fits | 35.9–71.8 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 2.87 GiB | fits | 64.8–129.6 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 2.87 GiB | fits | 64.8–129.6 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 2.87 GiB | fits | 64.8–129.6 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 2.87 GiB | fits | 97.1–194.2 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 2.87 GiB | fits | 129.6–259.3 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 2.87 GiB | fits | 129.6–259.3 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 2.87 GiB | fits | 253.4–506.7 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 2.87 GiB | fits | 253.4–506.7 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 2.87 GiB | fits | 253.4–506.7 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 2.87 GiB | fits | 32.3–64.6 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 2.87 GiB | fits | 64.8–129.6 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 2.87 GiB | fits | 129.6–259.3 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 2.87 GiB | fits | 25.3–50.7 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 2.87 GiB | fits | 25.3–50.7 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 2.87 GiB | fits | 25.3–50.7 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 2.87 GiB | fits | 57.6–115.3 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 2.87 GiB | fits | 57.6–115.3 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 2.87 GiB | fits | 57.6–115.3 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 2.87 GiB | fits | 86.6–173.1 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 2.87 GiB | fits | 115.3–230.6 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 2.87 GiB | fits | 115.3–230.6 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 2.87 GiB | fits | 172.9–345.8 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 2.87 GiB | fits | 172.9–345.8 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 2.87 GiB | fits | 76.0–152.0 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 2.87 GiB | fits | 197.6–395.2 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 2.87 GiB | fits | 60.8–121.6 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 2.87 GiB | fits | 106.4–212.8 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 2.87 GiB | fits | 151.3–302.7 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 2.87 GiB | fits | 212.8–425.6 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 2.87 GiB | fits | 94.6–189.2 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 2.87 GiB | fits | 141.9–283.8 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 2.87 GiB | fits | 189.2–378.4 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 2.87 GiB | fits | 202.7–405.4 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 2.87 GiB | fits | 378.4–756.7 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 2.87 GiB | fits | 378.4–756.7 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 2.87 GiB | fits | 202.7–405.4 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 2.87 GiB | fits | 96.3–192.6 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 2.87 GiB | fits | 54.1–108.1 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 2.87 GiB | fits | 57.6–115.3 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 2.87 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 2.87 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 | 4.43 GiB | fits | 17.2–34.4 estimated tokens/s |
| Mac mini M6 (12-core GPU) 24GB | 18.0 GiB | 4.43 GiB | fits | 19.1–38.2 estimated tokens/s |
| Mac mini M6 (12-core GPU) 32GB | 24.0 GiB | 4.43 GiB | fits | 19.1–38.2 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 24GB | 18.0 GiB | 4.43 GiB | fits | 34.5–69.0 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 48GB | 36.0 GiB | 4.43 GiB | fits | 34.5–69.0 estimated tokens/s |
| Mac mini M5 Pro (15-core CPU / 16-core GPU) 64GB | 48.0 GiB | 4.43 GiB | fits | 34.5–69.0 estimated tokens/s |
| Mac Studio M5 Max (32-core GPU) 36GB | 27.0 GiB | 4.43 GiB | fits | 51.7–103.4 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 64GB | 48.0 GiB | 4.43 GiB | fits | 69.0–138.0 estimated tokens/s |
| Mac Studio M5 Max (40-core GPU) 128GB | 96.0 GiB | 4.43 GiB | fits | 69.0–138.0 estimated tokens/s |
| Mac Studio M5 Ultra (30-core CPU / 64-core GPU) 96GB | 72.0 GiB | 4.43 GiB | fits | 134.8–269.6 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 256GB | 192.0 GiB | 4.43 GiB | fits | 134.8–269.6 estimated tokens/s |
| Mac Studio M5 Ultra (36-core CPU / 80-core GPU) 512GB | 384.0 GiB | 4.43 GiB | fits | 134.8–269.6 estimated tokens/s |
| MacBook Pro 14-inch M5 (10-core GPU) 24GB | 18.0 GiB | 4.43 GiB | fits | 17.2–34.4 estimated tokens/s |
| MacBook Pro 14-inch M5 Pro (20-core GPU) 48GB | 36.0 GiB | 4.43 GiB | fits | 34.5–69.0 estimated tokens/s |
| MacBook Pro 16-inch M5 Max (40-core GPU) 128GB | 96.0 GiB | 4.43 GiB | fits | 69.0–138.0 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 16GB | 12.0 GiB | 4.43 GiB | fits | 13.5–27.0 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 24GB | 18.0 GiB | 4.43 GiB | fits | 13.5–27.0 estimated tokens/s |
| Mac mini (2024) M4 (10-core GPU) 32GB | 24.0 GiB | 4.43 GiB | fits | 13.5–27.0 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 24GB | 18.0 GiB | 4.43 GiB | fits | 30.7–61.3 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 48GB | 36.0 GiB | 4.43 GiB | fits | 30.7–61.3 estimated tokens/s |
| Mac mini (2024) M4 Pro (16-core GPU) 64GB | 48.0 GiB | 4.43 GiB | fits | 30.7–61.3 estimated tokens/s |
| Mac Studio (2025) M4 Max (32-core GPU) 36GB | 27.0 GiB | 4.43 GiB | fits | 46.1–92.1 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 64GB | 48.0 GiB | 4.43 GiB | fits | 61.3–122.7 estimated tokens/s |
| Mac Studio (2025) M4 Max (40-core GPU) 128GB | 96.0 GiB | 4.43 GiB | fits | 61.3–122.7 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (28-core CPU / 60-core GPU) 96GB | 72.0 GiB | 4.43 GiB | fits | 92.0–184.0 estimated tokens/s |
| Mac Studio (2025) M3 Ultra (32-core CPU / 80-core GPU) 256GB | 192.0 GiB | 4.43 GiB | fits | 92.0–184.0 estimated tokens/s |
| NVIDIA GeForce RTX 3060 12GB | 10.8 GiB | 4.43 GiB | fits | 40.4–80.9 estimated tokens/s |
| NVIDIA GeForce RTX 3090 24GB | 21.6 GiB | 4.43 GiB | fits | 105.2–210.3 estimated tokens/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 14.4 GiB | 4.43 GiB | fits | 32.4–64.7 estimated tokens/s |
| NVIDIA GeForce RTX 4070 12GB GDDR6X | 10.8 GiB | 4.43 GiB | fits | 56.6–113.2 estimated tokens/s |
| NVIDIA GeForce RTX 4080 16GB | 14.4 GiB | 4.43 GiB | fits | 80.5–161.1 estimated tokens/s |
| NVIDIA GeForce RTX 4090 24GB | 21.6 GiB | 4.43 GiB | fits | 113.2–226.5 estimated tokens/s |
| NVIDIA GeForce RTX 5060 Ti 16GB | 14.4 GiB | 4.43 GiB | fits | 50.3–100.7 estimated tokens/s |
| NVIDIA GeForce RTX 5070 12GB | 10.8 GiB | 4.43 GiB | fits | 75.5–151.0 estimated tokens/s |
| NVIDIA GeForce RTX 5070 Ti 16GB | 14.4 GiB | 4.43 GiB | fits | 100.7–201.3 estimated tokens/s |
| NVIDIA GeForce RTX 5080 16GB | 14.4 GiB | 4.43 GiB | fits | 107.9–215.7 estimated tokens/s |
| NVIDIA GeForce RTX 5090 32GB | 28.8 GiB | 4.43 GiB | fits | 201.3–402.6 estimated tokens/s |
| NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB | 86.4 GiB | 4.43 GiB | fits | 201.3–402.6 estimated tokens/s |
| NVIDIA RTX 6000 Ada Generation 48GB | 43.2 GiB | 4.43 GiB | fits | 107.9–215.7 estimated tokens/s |
| Intel Arc B580 12GB | 10.8 GiB | 4.43 GiB | fits | 51.2–102.5 estimated tokens/s |
| AMD Ryzen AI Halo, Ryzen AI Max+ 395, 128GB | 89.6 GiB | 4.43 GiB | fits | 28.8–57.5 estimated tokens/s |
| NVIDIA DGX Spark 128GB / 4TB | 96.0 GiB | 4.43 GiB | fits | 30.7–61.3 estimated tokens/s |
| 2 × GeForce RTX 3090 24GB (conditional 48GB aggregate) | 40.8 GiB | 4.43 GiB | Needs verification | Estimated speed unavailable |
| 2 × GeForce RTX 5090 32GB (conditional 64GB aggregate) | 54.4 GiB | 4.43 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. Pinned license file confirmed in repository listing.
- architecture: Official source · verified 2026-10-22. Classification from model_type and expert/state-space configuration fields (qwen2); 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.