Wësche(@WescheNex1q):GLM 5.3-753B vs Qwen3.8-27B Same Eiffel Tower prompt, both local -GLM-5.3 full (4× DGX Spark): 120.9K tokens -Qwen3.8-27B BF16 (Mac Studio): 78.5K tokens The 28× bigger model thought 54% longer and still finished in half the time. Look at the towers. Setup: GLM-5.3 = @u1tra_instinct EXL3 2.75bpw + TensorFold TP4 recipe on 4× DGX Spark (MTP drafts, 262K context). Qwen3.8-27B = official BF16 via MLX. Temp 0.7, thinking at max, no output cap. Both MP4s checked: 1200 frames, 20.000 s, 1080p60. Fixes logged in REPAIRS.diff; I can share the method and scripts.

2026.10.04 18:34

GLM 5.3-753B vs Qwen3.8-27B Same Eiffel Tower prompt, both local -GLM-5.3 full (4× DGX Spark): 120.9K tokens -Qwen3.8-27B BF16 (Mac Studio): 78.5K tokens The 28× bigger model thought 54% longer and still finished in half the time. Look at the towers. Setup: GLM-5.3 = @u1tra_instinct EXL3 2.75bpw + TensorFold TP4 recipe on 4× DGX Spark (MTP drafts, 262K context). Qwen3.8-27B = official BF16 via MLX. Temp 0.7, thinking at max, no output cap. Both MP4s checked: 1200 frames, 20.000 s, 1080p60. Fixes logged in REPAIRS.diff; I can share the method and scripts.

显示更多

Tony@EnvolDev

2026.10.03 18:14

🚨 Opus 5.5: Today vs. At Launch — Nerfed? 👀 With everyone talking about a possible nerf, I had to test it myself! I reran the Eiffel Tower benchmark, and honestly, I prefer the launch result. It honestly feels like Opus 5.5 got little nerfed. The tower itself might have slightly better detail now, but the surrounding buildings, trees, river, and overall scene look noticeably worse to me. One run doesn’t prove a nerf, but this comparison surprised me. I want to see whether the same pattern shows up in the night train benchmark next! 🚂🌙 What do you think: nerf or normal variation? 🤔

显示更多