Nvidia just dropped their NVFP4 quant of Qwen3.8 Flash. Calibration dataset includes cnn_dailymail + Nemotron-Post-Training-Dataset-v2.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Where is Qwen 3.8 NVFP4? | 0 | 286 | August 17, 2026 | |
| Qwen3.8-Flash-Next 176B Now Available | 0 | 655 | August 26, 2026 | |
| Qwen3.8-Flash-Next NVFP4 on 2x DGX Spark - recipe, benchmarks, and the patch it needs to load | 4 | 833 | September 7, 2026 | |
| Qwen3.8-27B on DGX Spark using vllm: NVFP4 vs FP8 performance | 8 | 5479 | August 24, 2026 | |
| Faux NVFP2 Qwen3.8-Flash-Next Single Spark | 5 | 798 | September 18, 2026 | |
| Qwen3.8-Flash-Next on 1, 2 and 4 DGX Sparks with NVIDIA's official NVFP4 quant: 64 tok/s peak single stream | 27 | 6211 | September 10, 2026 | |
| Qwen3.8-Flash-Next NVFP4 vs GLM5.3-Flash NVFP4 | 4 | 2762 | August 31, 2026 | |
| Qwen 3.8-Flash-Next Single Node VLLM 24 tok/s | 0 | 1552 | August 28, 2026 | |
| DGX Spark, Nemotron3, and NVFP4: Getting to 65+ tps | 14 | 2598 | December 22, 2025 | |
| NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4 | 89 | 11928 | March 31, 2026 |