Qwen 3 8B
by Alibaba · qwen-3 family
8B
parameters
text-generation code-generation reasoning multilingual math tool-use summarization
Qwen 3 8B is the workhorse of the Qwen 3 dense lineup, offering an excellent balance of capability and resource efficiency. Features hybrid thinking mode for adaptive reasoning depth and supports tool calling for agentic workflows. At Q4 it fits on 8 GB GPUs with some headroom, and runs comfortably on 12-16 GB hardware. Strong at coding, math, and multilingual tasks — a direct upgrade over Llama 3.1 8B in most benchmarks.
Quick Start with Ollama
ollama run 8b-q4_K_M | Creator | Alibaba |
| Parameters | 8B |
| Architecture | transformer-decoder |
| Context | 128K tokens |
| Released | Apr 29, 2025 |
| License | Apache 2.0 |
| Ollama | qwen3:8b |
Quantization Options
| Format | File Size | VRAM Required | Quality | Ollama Tag |
|---|---|---|---|---|
| Q4_K_M rec | 5.2 GB | 7.5 GB |
★
★
★
★
★
| 8b-q4_K_M |
| Q8_0 | 8.9 GB | 11.5 GB |
★
★
★
★
★
| 8b-q8_0 |
| F16 | 16.5 GB | 20 GB |
★
★
★
★
★
| 8b-fp16 |
Compatible Hardware
for Q4_K_M (7.5 GB VRAM)
Compatible Hardware
| Hardware | VRAM | Type | Fit | Est. Speed |
|---|---|---|---|---|
| Mac Studio M4 Ultra 512GB | 512 GB | mac | Runs | ~109 tok/s |
| Mac Pro M2 Ultra 192GB | 192 GB | mac | Runs | ~107 tok/s |
| Mac Studio M4 Ultra 192GB | 192 GB | mac | Runs | ~109 tok/s |
| Mac Studio M4 Max 128GB | 128 GB | mac | Runs | ~73 tok/s |
| MacBook Pro M4 Max 128GB | 128 GB | mac | Runs | ~73 tok/s |
| MacBook Pro M3 Max 96GB | 96 GB | mac | Runs | ~53 tok/s |
| Mac mini M4 Pro 64GB | 64 GB | mac | Runs | ~36 tok/s |
| Mac Studio M4 Max 64GB | 64 GB | mac | Runs | ~73 tok/s |
| MacBook Pro M4 Max 64GB | 64 GB | mac | Runs | ~73 tok/s |
| Mac mini M4 Pro 48GB | 48 GB | mac | Runs | ~36 tok/s |
| MacBook Pro M3 Max 48GB | 48 GB | mac | Runs | ~53 tok/s |
| MacBook Pro M4 Max 48GB | 48 GB | mac | Runs | ~73 tok/s |
| MacBook Pro M4 Pro 48GB | 48 GB | mac | Runs | ~36 tok/s |
| Mac Studio M4 Max 36GB | 36 GB | mac | Runs | ~73 tok/s |
| MacBook Pro M3 Pro 36GB | 36 GB | mac | Runs | ~20 tok/s |
| NVIDIA GeForce RTX 5090 | 32 GB | gpu | Runs | ~239 tok/s |
| iMac M4 32GB | 32 GB | mac | Runs | ~16 tok/s |
| Mac mini M4 32GB | 32 GB | mac | Runs | ~16 tok/s |
| MacBook Air M4 32GB | 32 GB | mac | Runs | ~16 tok/s |
| AMD Radeon RX 7900 XTX | 24 GB | gpu | Runs | ~128 tok/s |
| NVIDIA GeForce RTX 3090 | 24 GB | gpu | Runs | ~125 tok/s |
| NVIDIA GeForce RTX 4090 | 24 GB | gpu | Runs | ~134 tok/s |
| iMac M3 24GB | 24 GB | mac | Runs | ~13 tok/s |
| Mac mini M2 24GB | 24 GB | mac | Runs | ~13 tok/s |
| Mac mini M4 Pro 24GB | 24 GB | mac | Runs | ~36 tok/s |
| MacBook Air M2 24GB | 24 GB | mac | Runs | ~13 tok/s |
| MacBook Air M4 24GB | 24 GB | mac | Runs | ~16 tok/s |
| MacBook Pro M4 Pro 24GB | 24 GB | mac | Runs | ~36 tok/s |
| AMD Radeon RX 7900 XT | 20 GB | gpu | Runs | ~107 tok/s |
| MacBook Pro M3 Pro 18GB | 18 GB | mac | Runs | ~20 tok/s |
| AMD Radeon RX 6800 XT | 16 GB | gpu | Runs | ~68 tok/s |
| AMD Radeon RX 7800 XT | 16 GB | gpu | Runs | ~83 tok/s |
| Intel Arc A770 | 16 GB | gpu | Runs | ~75 tok/s |
| NVIDIA GeForce RTX 4060 Ti 16GB | 16 GB | gpu | Runs | ~38 tok/s |
| NVIDIA GeForce RTX 4070 Ti Super | 16 GB | gpu | Runs | ~90 tok/s |
| NVIDIA GeForce RTX 4080 Super | 16 GB | gpu | Runs | ~98 tok/s |
| NVIDIA GeForce RTX 4080 | 16 GB | gpu | Runs | ~96 tok/s |
| NVIDIA GeForce RTX 5070 Ti | 16 GB | gpu | Runs | ~119 tok/s |
| NVIDIA GeForce RTX 5080 | 16 GB | gpu | Runs | ~128 tok/s |
| iMac M1 16GB | 16 GB | mac | Runs | ~9 tok/s |
| iMac M4 16GB | 16 GB | mac | Runs | ~16 tok/s |
| Mac mini M1 16GB | 16 GB | mac | Runs | ~9 tok/s |
| Mac mini M4 16GB | 16 GB | mac | Runs | ~16 tok/s |
| MacBook Air M2 16GB | 16 GB | mac | Runs | ~13 tok/s |
| MacBook Air M3 16GB | 16 GB | mac | Runs | ~13 tok/s |
| MacBook Air M4 16GB | 16 GB | mac | Runs | ~16 tok/s |
| MacBook Pro M1 16GB | 16 GB | mac | Runs | ~9 tok/s |
| MacBook Pro M2 Pro 16GB | 16 GB | mac | Runs | ~27 tok/s |
| AMD Radeon RX 7700 XT | 12 GB | gpu | Runs | ~58 tok/s |
| NVIDIA GeForce RTX 3060 12GB | 12 GB | gpu | Runs | ~48 tok/s |
| NVIDIA GeForce RTX 3080 12GB | 12 GB | gpu | Runs | ~122 tok/s |
| NVIDIA GeForce RTX 4070 Super | 12 GB | gpu | Runs | ~67 tok/s |
| NVIDIA GeForce RTX 4070 Ti | 12 GB | gpu | Runs | ~67 tok/s |
| NVIDIA GeForce RTX 4070 | 12 GB | gpu | Runs | ~67 tok/s |
| NVIDIA GeForce RTX 5070 | 12 GB | gpu | Runs | ~90 tok/s |
| NVIDIA GeForce RTX 2080 Ti | 11 GB | gpu | Runs | ~82 tok/s |
| NVIDIA GeForce RTX 3080 10GB | 10 GB | gpu | Runs | ~101 tok/s |
| AMD Radeon RX 7600 | 8 GB | gpu | Runs (tight) | ~38 tok/s |
| Intel Arc A750 | 8 GB | gpu | Runs (tight) | ~68 tok/s |
| NVIDIA GeForce RTX 3060 Ti | 8 GB | gpu | Runs (tight) | ~60 tok/s |
| NVIDIA GeForce RTX 3070 | 8 GB | gpu | Runs (tight) | ~60 tok/s |
| NVIDIA GeForce RTX 4060 Ti 8GB | 8 GB | gpu | Runs (tight) | ~38 tok/s |
| NVIDIA GeForce RTX 4060 | 8 GB | gpu | Runs (tight) | ~36 tok/s |
| MacBook Air M1 8GB | 8 GB | mac | Runs (tight) | ~9 tok/s |
| MacBook Air M2 8GB | 8 GB | mac | Runs (tight) | ~13 tok/s |
Benchmark Scores
73.5
mmlu