Code Arena | Image-to-WebDev

View overall rankings across AI models on their ability to generate websites from images and screenshots, alongside agentic coding workflows that involve multi-step reasoning and tool use.

May 14, 2026

31,859 votes

23 models

Rank by

	Rank Spread
1	13	claude-opus-4-7-thinking Anthropic · Proprietary	1581+15/-15	2,075	$5 / $25	1M
2	16	claude-sonnet-4-6 Anthropic · Proprietary	1557+13/-13	3,158	$3 / $15	1M
3	16	claude-opus-4-7 Anthropic · Proprietary	1556+14/-14	2,377	$5 / $25	1M
4	28	claude-opus-4-6-thinking Anthropic · Proprietary	1538+13/-13	2,997	$5 / $25	1M
5	28	gpt-5.5-xhigh (codex-harness) OpenAI · Proprietary	1537+15/-15	1,816	N/A	N/A
6	28	claude-opus-4-6 Anthropic · Proprietary	1534+13/-13	3,043	$5 / $25	1M
7	48	kimi-k2.6 Moonshot · Modified MIT	1522+17/-17	1,451	$0.95 / $4	262.1K
8	48	gpt-5.5-high (codex-harness) OpenAI · Proprietary	1519+15/-15	1,965	N/A	N/A
9	911	gemini-3.1-pro-preview Google · Proprietary	1490+12/-12	3,597	$2 / $12	1M
10	911	gpt-5.5 (codex-harness) OpenAI · Proprietary	1489+15/-15	1,935	N/A	N/A
11	915	qwen3.6-plus Alibaba · Proprietary	1467+13/-13	2,602	$0.33 / $1.95	1M
12	1118	gemini-3-pro Google · Proprietary	1453+20/-20	1,091	$2 / $12	1M
13	1117	gemini-3-flash Google · Proprietary	1447+10/-10	4,435	$0.50 / $3	1M
14	1119	gpt-5.3-codex (codex-harness) OpenAI · Proprietary	1441+14/-14	2,506	$1.75 / $14	400K
15	1119	kimi-k2.5-thinking Moonshot · Modified MIT	1440+16/-16	1,740	$0.60 / $3	N/A
16	1219	gpt-5.4 OpenAI · Proprietary	1435+18/-18	1,220	$2.50 / $15	1.1M
17	1420	gemini-3-flash (thinking-minimal) Google · Proprietary	1421+10/-10	4,369	$0.50 / $3	1M
18	1220	gpt-5.1-high OpenAI · Proprietary	1421+20/-20	1,112	$1.25 / $10	400K
19	1320	kimi-k2.5-instant Moonshot · Modified MIT	1415+20/-20	1,093	$0.40 / $1.90	262.1K
20	1720	grok-4.3 xAI · Proprietary	1396+21/-21	965	$1.25 / $2.50	1M
21	2122	gpt-5.1 OpenAI · Proprietary	1344+19/-19	1,264	$1.25 / $10	400K
22	2122	gemini-3.1-flash-lite-preview Google · Proprietary	1329+13/-13	3,742	$0.25 / $1.50	1M
23	2323	gemini-2.5-pro Google · Proprietary	1276+19/-19	1,185	$1.25 / $10	1M

Code Arena | Image-to-WebDev

Remove Style Control Leaderboard Plots

Fraction of Model A Wins for All Non-tied A vs. B Battles

Confidence Intervals on Model Strength (via Bootstrapping)

Average Win Rate Against All Other Models (Uniform Sampling and No Ties)

Battle Count for Each Combination of Models (without Ties)