Qwen3.5
Qwen3.5-397B-A17B
Alibaba (Qwen team) · open-permissive apache-2.0
| Parameters (B) | 403.398 |
|---|---|
| Architecture | moe |
| Layers / KV heads / head dim | 60 / 2 / 256 |
| Context window | 262144 |
| Modalities | in: text, image; out: text |
| GGUF | available |
| Commercial use | yes |
Minimum hardware (estimate)
At Q4_K_M and 8k context, total working set ≈ 246.5 GiB. On a machine with 64 GiB system RAM and no discrete GPU: Won't run. Limiting factor: No GPU and system RAM insufficient for weights + KV.
Scores
| Benchmark | Value | Uncertainty | Run by | Source |
|---|---|---|---|---|
| GPQA Diamond | 86.4% | ±0.0245 | third-party | epoch-ai |
| OTIS Mock AIME 2024-2025 | 88.9% | ±0.0474 | third-party | epoch-ai |
Provenance
- identity, license, architecture, parameter count: https://huggingface.co/Qwen/Qwen3.5-397B-A17B (retrieved 2026-08-10)
HF: Qwen/Qwen3.5-397B-A17B · snapshot 2026-08-10