Qwen3-8B
Qwen3-8B is an open-weights language model from Qwen, released April 28, 2025. It has 8.2B parameters (dense). The context window is 128K tokens. On the Epoch Capabilities Index it scores 132.1, rank 152 of 256 rated models. API price: $0.18 input / $0.7 output per million tokens.
Check it against your own PC →
Can you run Qwen3-8B locally?
The smallest common GPU it runs on is 6 GB of VRAM (with 32 GB of RAM): Q6_K, split GPU / CPU, roughly 6 tokens per second. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Runs, slowly | Q6_K (≈ lossless) | 5.2 GB | 4.5 GB | ~6 |
| 8 GB VRAM | Runs on GPU | Q3_K_M (okay) | 7.0 GB | – | ~39 |
| 12 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 9.7 GB | – | ~31 |
| 16 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 9.7 GB | – | ~43 |
| 24 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 9.7 GB | – | ~77 |
| 32 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 9.7 GB | – | ~140 |
Ready-made quantizations on Hugging Face: Qwen/Qwen3-Embedding-8B-GGUF · mradermacher/Qwen3-8B-Base-GGUF · Mungert/Qwen3-Reranker-8B-GGUF · QuantFactory/Qwen3-Reranker-8B-GGUF · mradermacher/Qwen3-8B-Base-i1-GGUF · QuantFactory/Qwen3-8B-GGUF
Benchmark results
| Benchmark (Epoch AI) | Score |
|---|---|
| OTIS Mock AIME | 56.1% |
| GPQA Diamond | 44.1% |
Links
Hugging Face: Qwen/Qwen3-8B-Base · Epoch AI benchmarks
Models with a similar score
Gemini 2.0 Pro · o1-mini · DeepSeek-V3 (Mar 2025) · Qwen3-30B-A3B-Instruct (Jul 2025) · gpt-oss-20b · GPT-4.1 mini · Qwen3-30B-A3B · Gemini 2.0 Flash (Feb 2025)
More from Qwen
Qwen3.8 Omni Flash · Qwen3.8 Max (0902) · Qwen3.8 Flash · Qwen3.8 Flash Next · Qwen 3.8 27B · Qwen3.8-2.4T-A95B · Qwen 3.8 Max · Qwen3.7 Flash