Llama 3.2 3B Instruct Legacy
Files and memory at 8K context
No verified quantized file for the requested formats.
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. Legacy model: sizes on device pages are estimated from parameter count.
Quality
Artificial Analysis Intelligence Index
Unknown
Arena text rating
Unknown
Scores describe the evaluated hosted variant, not each quantized file. AA and Arena are never averaged.
Reported speeds 23
| Hardware | Quant | Backend | Prompt t/s | Gen t/s | Date | Source |
|---|---|---|---|---|---|---|
| M4, 10 GPU cores (Mac mini) | Q4_K_M | llama.cpp Metal | 530 | 46 | 2026-05-12 | llama.cpp #4167 |
| Mac Studio M1 Ultra (64 core GPU, 128GB) | Q4_K_M | llama.cpp Metal | 2,143 | 128 | 2026-02-07 | ai-benchmarks #48 |
| Mac Studio M1 Ultra (64 core GPU, 128GB) | Q4_K_M | llama.cpp Metal | 1,912 | — | 2026-02-07 | ai-benchmarks #48 |
| NVIDIA GeForce RTX 4070 Ti (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 11,979 | 196 | 2025-12-10 | ai-benchmarks #45 |
| NVIDIA GeForce RTX 4070 Ti (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 6,542 | — | 2025-12-10 | ai-benchmarks #45 |
| NVIDIA GeForce RTX 4090 (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 20,323 | 335 | 2025-12-09 | ai-benchmarks #43 |
| NVIDIA GeForce RTX 4090 (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 12,746 | — | 2025-12-09 | ai-benchmarks #43 |
| NVIDIA GeForce RTX 3060 12GB (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 4,006 | 123 | 2025-12-06 | ai-benchmarks #40 |
| NVIDIA GeForce RTX 3060 12GB (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 2,801 | — | 2025-12-06 | ai-benchmarks #40 |
| AMD Radeon AI Pro R9700 32GB (Intel Core Ultra 7 265K system) | Q4_K_M | llama.cpp Vulkan | 4,051 | 196 | 2025-11-25 | ai-benchmarks #35 |
| AMD Radeon AI Pro R9700 32GB (Intel Core Ultra 7 265K system) | Q4_K_M | llama.cpp Vulkan | 3,570 | — | 2025-11-25 | ai-benchmarks #35 |
| Framework Desktop Mainboard (AMD Ryzen AI Max+ 395, 128 GB) | Q4_K_M | llama.cpp Vulkan | 1,581 | 88.1 | 2025-08-07 | ai-benchmarks #21 |
| Framework Desktop Mainboard (AMD Ryzen AI Max+ 395, 128 GB) | Q4_K_M | llama.cpp Vulkan | 1,060 | — | 2025-08-07 | ai-benchmarks #21 |
| Framework Desktop Mainboard (AMD Ryzen AI Max+ 395, 128 GB) | Q4_K_M | llama.cpp ROCm | 574 | 75 | 2025-08-07 | ai-benchmarks #21 |
| NVIDIA GeForce RTX 2080 Ti (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 4,418 | 180 | 2025-12-18 | ai-benchmarks #46 |
| NVIDIA GeForce RTX 2080 Ti (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 3,371 | — | 2025-12-18 | ai-benchmarks #46 |
| Mac Studio M3 Ultra 512GB | Q4_K_M | llama.cpp Metal | 3,334 | 155 | 2025-12-08 | ai-benchmarks #42 |
| Mac Studio M3 Ultra 512GB | Q4_K_M | llama.cpp Metal | 2,915 | — | 2025-12-08 | ai-benchmarks #42 |
| NVIDIA RTX A4000 (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 5,540 | 149 | 2025-11-28 | ai-benchmarks #36 |
| NVIDIA RTX A4000 (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 3,779 | — | 2025-11-28 | ai-benchmarks #36 |
| NVIDIA GeForce RTX 3080 Ti 12GB (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 7,371 | 263 | 2025-11-28 | ai-benchmarks #39 |
| NVIDIA GeForce RTX 3080 Ti 12GB (Intel Core Ultra 265K custom PC) | Q4_K_M | llama.cpp Vulkan | 6,476 | — | 2025-11-28 | ai-benchmarks #39 |
| NVIDIA Tegra X1 | Q4_1 | llama.cpp Vulkan | 40.8 | 4.9 | 2025-01-12 | llama.cpp #10879 |
Reported by the linked sources, not reproduced by LocalFit. All rows in the speed database