~ same prompt, different LLM — every model gets the same brief in Blender, you judge the render ~
Google / Gemini 3.8 Flash / thinking:medium has entered the arena NEW!
LeaderboardGalleryVoteRunsTasks
RankContestantRatingVotesRunsDone %Success %Avg $Avg callsAvg time
1OpenAI / GPT 6 Astra / thinking:mediumcompare →17632010100$0.57—260s
2OpenAI / GPT 5.6 Lunacompare →1464211100100$0.03—67s
provGoogle / Gemini 3.5 Flash / thinking:mediumcompare →1273110100$0.16—145s

Model comparison

each point is a model (One-shot, corgi-beach-ball) — hover for exact values
GoogleOpenAI unrated
12001359151816771837$0.021$0.050$0.123$0.300$0.730Bradley–Terry ratingOpenAI / GPT 6 Astra / thinking:mediumOpenAI / GPT 5.6 LunaGoogle / Gemini 3.5 Flash / thinking:mediumMean cost per task (USD, log scale)

By dimension

lens rounds and chip verdicts feed these — none cast yet
Lighting
no data yet
Composition
no data yet
Modeling
no data yet
Materials
no data yet
Render
no data yet
Adherence
no data yet

Modes compared

each contestant's ratings across run modes (mixed ladder)
ContestantFullOne-shot3D-only
DeepSeek / Deepseek V4 Flash1502——
DeepSeek / Deepseek V4 Pro1486——
Google / Gemini 3.1 Flash Lite1273——
Google / Gemini 3.5 Flash / thinking:medium15791191—
Google / Gemma 4 31b It1356——
Z.AI / GLM 5.21394——
OpenAI / GPT 5.6 Luna17711382—
OpenAI / GPT 5.6 Luna / thinking:max1716——
OpenAI / GPT 5.6 Sol1834——
OpenAI / GPT 6 Astra / thinking:medium—1668—
Anthropic / Haiku 4.5 / thinking:xhigh1348——
Vote A/BGalleryRunsbradley-terry fit over 21 decisive votes
Votes cast on this bench: 001671 · 163 runs · 26 contestants
Best viewed with Netscape Navigator 4.07 at 800×600 · real runs, live ratings