Arena Image-to-WebDev · Open Analysis

#	Model	Effort	Score	Cost / task	Harness
1	Opus 5.5	max	1749	–	–
2	Sonnet 5.5	xhigh	1740	–	–
3	GPT-6 Astra	max	1731	–	–
4	Fable 5.1	max	1720	–	–
5	GPT-6.1 Sol	max	1703	–	–
6	GPT-6 Sol	max	1661	–	–
7	Opus 5	max	1661	–	–
8	Muse Spark 1.3	max	1642	–	–
9	Qwen 3.8 Max	max	1638	–	–
10	Fable 5	high	1622	–	–
11	DeepSeek V4.1 Flash	max	1609	–	–
12	GPT-5.6 Sol	xhigh	1608	–	codex-harness
13	Gemini 3.7 Flash	high	1603	–	–
14	Grok 4.6	high	1601	–	–
15	MiMo V2.6 Flash	high	1597	–	–
16	MiMo V2.6 Pro	high	1595	–	–
17	Gemini 3.8 Flash	high	1593	–	–
18	GLM 5.3 Flash	max	1581	–	–
19	Kimi K3	max	1579	–	–
20	Opus 4.7	high	1578	–	–
21	Grok 4.7	xhigh	1577	–	–
22	Qwen 3.8 27B	max	1576	–	–
23	GPT-6 Luna	max	1570	–	–
24	Grok 4.5	high	1564	–	–
25	Gemini 3.6 Flash	high	1545	–	–
26	Opus 4.6	high	1543	–	–
27	GPT-5.6 Terra	xhigh	1540	–	codex-harness
28	Muse Spark 1.1	xhigh	1540	–	–
29	Sonnet 5	high	1539	–	–
30	GPT-5.5	xhigh	1532	–	codex-harness
31	Sonnet 4.6	max	1529	–	–
32	Kimi K2.6	xhigh	1521	–	–
33	GPT-5.6 Luna	xhigh	1517	–	codex-harness
34	Kimi K2.7 Code	–	1513	–	–
35	Gemini 3.5 Flash	xhigh	1512	–	–
36	Gemini 3.1 Pro	high	1481	–	–
37	MiniMax M3	max	1473	–	–
38	Qwen 3.6 Plus	max	1472	–	–
39	Gemini 3 Flash	high	1451	–	–
40	Gemini 3.5 Flash-Lite	xhigh	1447	–	–
41	GPT-5.3 Codex	xhigh	1444	–	codex-harness
42	Gemini 3 Pro	high	1441	–	–
43	GPT-5.4	xhigh	1440	–	–
44	Kimi K2.5	high	1440	–	–
45	GLM 5 V	high	1425	–	–
46	GPT-5.1	high	1421	–	–
47	Inkling Small	xhigh	1410	–	–
48	Grok 4.3	xhigh	1375	–	–
49	Mistral Large 3	none	1362	–	–
50	Gemini 3.1 Flash-Lite	high	1337	–	–
51	Gemini 2.5 Pro	high	1285	–	–