[Performance Regression] Significant drop in prompt evaluation speed and increased load time in `0.30.0-RC15` on qwen3:0.6b
![[Performance Regression] Significant drop in prompt evaluation speed and increased load time in `0.30.0-RC15` on qwen3:0.6b](https://www.chat-gpts.plus/wp-content/uploads/2026/07/16148-f90cd745-768x403.jpg)
用户在 Debian 13 系统上,使用 NVIDIA RTX 3090(24GB VRAM)显卡,从 Ollama 0.23.4 升级到 0.30.0-RC15 后,运行 qwen3:0.6b 小模型时发现 prompt 评估速度从 1326.78 tokens/s 下降到 322.22 toke








