Atlas

Benchmarks

← All benchmarks

Arena Code WebDev — Consumer Product

Code

Arena Code WebDev rating for consumer-product and platform applications. Anonymous generated applications are compared through blind pairwise user votes.

Top models (higher is better)

ModelScore
Claude Opus 51651
Kimi K31615
Claude Fable 51562
GLM-5.21558
DeepSeek-V4-Flash1553
Opus 4.81549
Opus 4.71541
Opus 4.61533
Muse Spark 1.11531
Grok 4.51524
Sonnet 4.61522
Sonnet 51516
GLM-5.11514
Kimi K2.61507
Gemini 3.6 Flash1494
MiniMax M31481
Hy31477
Gemini 3.5 Flash1472
Kimi K2.7 Code1470
MiMo-V2.5-Pro1462
Opus 4.51460
Qwen3.6 Plus (2026-04-02)1451
MiMo-V2.51451
DeepSeek-V4-Pro1448
Gemini 3 Flash Preview1445
GLM-51444
MiMo-V2-Pro1443
Kimi K2.51439
Gemini 3 Pro Preview1436
GLM-4.71435
GLM 5V Turbo1428
Inkling1426
Gemini 3.1 Pro Preview1424
GPT-51420
GPT-5.21418
MiniMax M2.71407
GPT-5.3-Codex1403
GPT-5.11399
Qwen3.5 397B A17B1397
GPT-5.41396
Loading Atlas data…