Skip to main content
AI Choice Engine
← Benchmark sources

arena.ai · benchmark source

Arena

Live human-preference and agent boards from Arena (LMArena). Scores are board-specific — Text Arena Score, WebDev Arena Score, and Agent net improvement are not interchangeable. The Arena product home at arena.ai/ and the long-running lmarena.ai leaderboards are the same operator (verified 2026-08-29 via arena.ai root and the LMArena research pages; see https://lmarena.ai/ for the historic domain).

Open arena.ai3 boards on this publisher

Web development

Code Arena · WebDev

A live human-preference leaderboard for models that turn natural-language prompts into working front-end web applications, scored with Arena’s Bradley-Terry Arena Score.

Why it matters

It captures the prompt-to-product loop — layout, interactions, and whether the app feels usable — rather than only repository issue repair.

Use it when

You are choosing a model for UI/app generation and want a live preference signal alongside coding-agent benchmarks.

The limitation

Votes reflect voter taste on sandboxed app demos. Arena Score is board-specific and not comparable to Agent net improvement, SWE-bench, or DeepSWE percentages.

Leaderboard snapshot

Code Arena · WebDev results in a readable view

Use the tabs for the primary score and other published fields from this board. The source link remains authoritative for the live table.

Arena WebDev capture (138 ranked rows, board dated 2026-10-01). Top: claude-opus-5.5-max (1815), gpt-6-astra-max (1788), claude-sonnet-5.5-xhigh (1786).

Leader among verified rows

claude-opus-5.5-max

1,815

Rows shown

106

Arena Score

Source / snapshot date

2026-10-01

25 days old · not a live API feed

Score profile

Arena Score by published row

Higher is better in this view

claude-opus-5.5-max
1,815
gpt-6-astra-max
1,788
claude-sonnet-5.5-xhigh
1,786
gpt-6.1-sol-max
1,758
claude-fable-5.1-max
1,749
claude-sonnet-5.5-high
1,715
claude-opus-5-max
1,695
gpt-6-sol-max
1,689
gemini-4-argon-high
1,680
qwen3.8-max
1,671
qwen3.8-max-0902
1,670
claude-opus-5-high
1,660
kimi-k3-max
1,658
muse-spark-1.3-max
1,657
grok-4.7-xhigh
1,638
qwen3.8-flash-next
1,637
hy4-preview
1,633
claude-fable-5-high
1,625
muse-spark-1.3 (xHigh)
1,624
glm-5.3-max
1,623
deepseek-v4.1-flash-max
1,620
gpt-5.6-sol-xhigh (codex-harness)
1,620
grok-4.6-high
1,620
mimo-v2.6-pro
1,618
glm-5.3-flash
1,616
glm-5.2-max
1,605
gemini-3.7-flash-high
1,592
qwen3.8-27b
1,591
deepseek-v4-pro-high-20260813
1,583
gemini-3.8-flash-high
1,583
deepseek-v4-flash-high
1,581
gpt-6-luna-max
1,579
step-5-preview-high
1,570
claude-opus-4-7-high
1,557
claude-opus-4-8-high
1,556
claude-opus-4-7
1,555
grok-4.5
1,552
claude-opus-4-6-high
1,546
muse-spark-1.1
1,542
claude-sonnet-5-high
1,539
claude-opus-4-6
1,537
gemini-3.6-flash-high
1,536
claude-opus-4-8
1,534
muse-spark-1.2 (xHigh)
1,532
claude-sonnet-4-6
1,521
gpt-5.6-terra-xhigh (codex-harness)
1,519
gpt-5.6-luna-xhigh (codex-harness)
1,518
seed-2.1-pro-preview
1,517
qwen3.7-max-20260517
1,515
gpt-5.5-xhigh (codex-harness)
1,512
kimi-k2.6
1,509
glm-5.1
1,508
hy3
1,507
gemini-3.5-flash-high
1,499
claude-opus-4-5-20251101-high-32k
1,493
gemini-3.5-flash-medium
1,490
gpt-5.5-high (codex-harness)
1,487
minimax-m3
1,482
qwen3.6-max-preview
1,482
mimo-v2.5-pro
1,478
kimi-k2.7-code
1,473
claude-opus-4-5-20251101
1,468
gpt-5.4-high (codex-harness)
1,465
deepseek-v4-pro-high-preview
1,463
qwen3.6-plus
1,461
gpt-5.5 (codex-harness)
1,456
deepseek-v4-pro
1,446
gemini-3.1-pro-preview
1,446
gpt-5.4-medium (codex-harness)
1,444
gemini-3.5-flash-lite
1,440
gemini-3-flash
1,439
gemini-3-pro
1,439
mimo-v2.5
1,438
kimi-k2.5-thinking
1,436
glm-4.7
1,434
glm-5
1,434
mimo-v2-pro
1,433
deepseek-v4-flash-high-preview
1,430
gpt-5-medium
1,417
gpt-5.2
1,415
inkling
1,413
gpt-5.3-codex (codex-harness)
1,408
Inkling Small
1,408
kimi-k2.5-instant
1,404
glm-5v-turbo
1,400
qwen3.5-397b-a17b
1,399
gpt-5.4-mini-high
1,397
minimax-m2.7
1,397
gpt-5.1-medium
1,395
gpt-5.4
1,395
claude-sonnet-4-5-20250929-high-32k
1,393
claude-opus-4-1-20250805
1,389
minimax-m2.5
1,386
claude-sonnet-4-5-20250929
1,384
gemini-3-flash (thinking-minimal)
1,384
minimax-m2.1-preview
1,384
grok-4.20-beta-0309-reasoning
1,374
gpt-5.3-codex (codex-harness)
1,371
solar-pro4
1,370
gemma-4-31b
1,365
gemma-4-26b-a4b
1,358
claude-haiku-4-5-20251001
1,329
mistral-medium-3.5
1,263
gemini-3.1-flash-lite-preview
1,256
mistral-large-3
1,230
gemini-2.5-pro
1,227

Full Code Arena WebDev overall table captured 2026-10-01 from arena.ai/leaderboard/code/webdev (138 models, 838,388 votes; board dated 2026-10-01). Arena Score is not a percentage and is not comparable to Agent net improvement or DeepSWE pass rates. Preliminary ratings are marked per row.

Open Code Arena · WebDev leaderboard

Showing the top 100 of 138 entries plus every model with a page in our catalog. 32 lower-ranked entries are on the full source table.

Evidence in this catalog

Where Code Arena · WebDev fits

Closest verified examples we currently carry. A missing score is not a zero.

Model examples

  • Published source observation captured 2026-10-01. Source-specific result; no independent rerun or universal capability verdict.

  • Published source observation captured 2026-10-01. Source-specific result; no independent rerun or universal capability verdict.

  • Published source observation captured 2026-10-01. Source-specific result; no independent rerun or universal capability verdict.

Harness examples

No directly comparable harness score is verified here yet.

Open this board on Code Arena · WebDev leaderboard