Blender Bench v1: GPT-6 Astra leads 3D benchmark, Claude Opus 5.5 wins modeling and cloth
yshan2u · x · 2026-10-01
refreshdev released Blender Bench v1 (BBv1), benchmarking frontier LLMs on 20 realistic 3D production workflows including animation, modeling, rigging, cloth and surfacing. GPT-6 Astra leads overall; Claude Opus 5.5 wins modeling and cloth simulation.
More from Models
- RLHF co-author Diogo Almeida discusses the philosophy behind TypeSafe's Jev model — AxSaucedo · 2026-10-01
- Early user: GPT-6 Sol feels slower and dumber than 5.6 in Codex — burkov · 2026-10-01
- GPT-6.1 Sol Plots 3-Day Earth-Mars Transfer Orbit at Constant 1/3 g — MikePFrank · 2026-10-01
- Codex Pro Users Report Credits Vanishing Before Included Usage Is Exhausted — Constant-Potential-9 · 2026-10-01
- Community benchmark: Mercury Decide hits 66.7% accuracy, trailing Jev's 83.3% — FlexDruk · 2026-10-01
- Qwen Flash Next MTP work resumes with official GGUF quants and llama.cpp PR — jacek2023 · 2026-10-01