Benchmark source guide
Design Arena Benchmark Results
Design Arena benchmark results used for web, interface, game, and data visualization rankings with source details.
Stored observations
976
Matched models
47
Latest observation
Jul 17, 2026
How it is used
Method inside Vibe Bench
Vibe Bench stores category results as a separate evidence family. These rankings do not change the main Vibe Coding Index.
Category ELOWin rateGeneration timeTournament volume
Read with care
Known limits
Results measure preference inside each Design Arena category and should not be read as a complete general coding score.
Every displayed result keeps its observed date and source status so preliminary or stale evidence remains visible.
Current category leaders
Top Design Arena results by category
3d
Claude Fable 51,368
GLM 5.21,367
Grok 4.51,350
Claude Opus 4.6 (Thinking)1,347
Kimi K2.61,346
asciiart
Claude Fable 51,376
Claude Opus 4.71,324
Gemini 3.1 Pro Preview1,313
Claude Opus 4.81,306
Gemini 3.5 Flash1,305
codecategories
GLM 5.21,348
Claude Fable 51,345
Claude Opus 4.61,333
Grok 4.51,330
Claude Opus 4.6 (Thinking)1,328
dataviz
Claude Fable 51,362
Muse Spark 1.11,324
Claude Opus 4.61,321
Claude Opus 4.71,319
GLM 5.11,318
gamedev
Claude Fable 51,396
GLM 5.21,349
Claude Sonnet 51,345
GPT-5.51,343
Claude Opus 4.6 (Thinking)1,337
graphicdesign
Riverflow 2.5 Pro1,469
Gemini 3.1 Flash Image Gen 2K (Nano Banana 2)1,288
Gemini 3 Pro Image Gen 2K (Nano Banana Pro)1,277
Gemini 3.1 Flash Image Gen (Nano Banana 2)1,276
Gemini 3 Pro Image Preview1,253
image
Riverflow 2.5 Pro1,406
Gemini 3.1 Flash Image Gen 2K (Nano Banana 2)1,299
Gemini 3.1 Flash Image Gen (Nano Banana 2)1,291
Gemini 3 Pro Image Gen 2K (Nano Banana Pro)1,266
Gemini 3 Pro Image Preview1,239
imageediting
Riverflow 2.5 Pro1,372
Gemini 3 Pro Image Preview 2K1,267
Gemini 3 Pro Image Preview1,260
MAI-Image-2.51,254
FLUX.2 Pro Edit1,149