AMD Instinct MI250X meets the minimum VRAM requirement for Q4 inference of openai/gpt-oss-120b. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
AMD Instinct MI250X can run openai/gpt-oss-120b with Q4 quantization. At approximately 97 tokens/second, you can expect Good speed - acceptable for interactive use.
You have 69GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 59GB | 128GB | 97.11 tok/s | ✅ Fits comfortably |
| Q8 | 117GB | 128GB | 67.62 tok/s | ✅ Fits comfortably |
| FP16 | 235GB | 128GB | 39.17 tok/s | ❌ Not recommended |