~ same prompt, different LLM — every model gets the same brief in Blender, you judge the render ~
HomeLeaderboardGallerySandboxTasksVote!RunsAbout
OpenAI / GPT 6 Astra / thinking:max has entered the arena NEW!
www.blender-bench.org   Ru:157
//runs/goldfish/
Scene
goldfish__gpt-6-astra-think-medium__r1__20260905-121532 renderAI cam
A goldfish · r1 · 2026-09-05
OpenAI / GPT 6 Astra / thinking:mediumNO DONE
+ COMPARE
goldfish__ox-alpha-think-low__r2__20260825-194432 renderAI cam
A goldfish · r2 · 2026-08-25
Ox Alpha / thinking:lowMODEL_ERROR
+ COMPARE
goldfish__ox-alpha-think-low__r1__20260825-184521 renderAI cam
A goldfish · r1 · 2026-08-25
Ox Alpha / thinking:lowMODEL_ERROR
+ COMPARE
goldfish__gpt-5.6-luna-think-max__r1__20260731-150409 renderAI cam
A goldfish · r1 · 2026-07-31
OpenAI / GPT 5.6 Luna / thinking:maxDONE ✓
✓ IN COMPARE
goldfish__deepseek-v4-flash-0731-think-max__r2__20260731-132248 renderauto
A goldfish · r2 · 2026-07-31
DeepSeek / Deepseek V4 Flash 0731 / thinking:maxNO DONE
+ COMPARE
goldfish__claude-opus-5-think-medium__r1__20260726-022536 renderAI cam
A goldfish · r1 · 2026-07-26
Anthropic / Opus 5 / thinking:mediumDONE ✓
+ COMPARE
goldfish__gpt-5.6-sol-think-max__r1__20260719-211615 renderAI cam
A goldfish · r1 · 2026-07-19
OpenAI / GPT 5.6 Sol / thinking:maxDONE ✓
+ COMPARE
goldfish__gemini-3.5-flash-think-medium__r1__20260719-185314 renderauto
A goldfish · r1 · 2026-07-19
Google / Gemini 3.5 Flash / thinking:mediumNO DONE
+ COMPARE
goldfish__deepseek-v4-flash__r1__20260719-182519 renderauto
A goldfish · r1 · 2026-07-19
DeepSeek / Deepseek V4 FlashNO DONE
+ COMPARE
goldfish__sonnet-5-think-low__r1__20260719-171547 renderauto
A goldfish · r1 · 2026-07-19
Anthropic / Sonnet 5 / thinking:lowNO DONE
+ COMPARE
goldfish__gpt-5.6-luna__r1__20260716-205630 renderAI cam
A goldfish · r1 · 2026-07-16
OpenAI / GPT 5.6 LunaNO DONE
+ COMPARE
goldfish__gpt-5.6-luna__r1__20260716-005004 renderAI cam
A goldfish · r1 · 2026-07-16
OpenAI / GPT 5.6 LunaDONE ✓
+ COMPARE
Compare tray — 1 of 2
A: goldfish · OpenAI / GPT 5.6 Luna / thinking:max · r1 ✕CLEARCOMPARE →
★ Source on GitHub SOON · ♥ Support the bench
Votes cast on this bench: 000504 · 157 runs · 25 contestants
Best viewed with Netscape Navigator 4.07 at 800×600 · real runs, live ratings