Vending-Bench 2 · Open Analysis

#	Model	Effort	Score	Cost / task	Harness
1	GPT-6 Astra	max	15514.70 USD	–	–
2	GPT-6 Sol	max	14427.85 USD	–	–
3	Gemini 4	max	13718.16 USD	–	–
4	Muse Wonder Full (Max)	max	12291.56 USD	–	–
5	Opus 5	max	11181.87 USD	–	–
6	Sonnet 5.5	max	11165.28 USD	–	–
7	Opus 4.7	max	10936.76 USD	–	–
8	Grok 4.7	max	10536.83 USD	–	–
9	GPT-5.6 Sol	max	9619.37 USD	–	–
10	Muse Wonder	max	9432.18 USD	–	–
11	Opus 5.5	max	9235.25 USD	–	–
12	Grok 4.6	max	9047.03 USD	–	–
13	GLM 5.2	max	8313.78 USD	–	–
14	GLM 5.3	max	8163.61 USD	–	–
15	Opus 4.6	max	8017.59 USD	–	–
16	GPT-5.5	xhigh	7523.84 USD	–	–
17	GPT-5.6 Terra	max	7343.21 USD	–	–
18	Sonnet 4.6	max	7204.14 USD	–	–
19	Muse Spark 1.1	max	6520.47 USD	–	–
20	Sonnet 5	max	6377.70 USD	–	–
21	Kimi K2.6	max	6204.57 USD	–	–
22	GPT-5.4	xhigh	6144.18 USD	–	–
23	GPT-5.3 Codex	xhigh	5940.12 USD	–	–
24	Opus 4.8	high	5787.43 USD	–	–
25	Fable 5	high	5680.26 USD	–	–
26	GLM 5.1	max	5634.41 USD	–	–
27	Gemini 3 Pro	high	5478.16 USD	–	–
28	Fable 5.1	max	5421.56 USD	–	–
29	Gemini 3.5 Flash	xhigh	5396.42 USD	–	–
30	Kimi K3	max	5165.04 USD	–	–
31	Qwen 3.6 Plus	max	5114.87 USD	–	–
32	Gemini 3.8 Flash	max	5093.79 USD	–	–
33	Kimi K2.7 Code	max	5082.94 USD	–	–
34	Opus 4.5	high	4967.06 USD	–	–
35	Grok 4.20	xhigh	4662.85 USD	–	–
36	GLM 5	high	4432.12 USD	–	–
37	Qwen 3.6 Max	max	4254.19 USD	–	–
38	GPT-5.6 Luna	max	4094.71 USD	–	–
39	Grok 4.5	max	3887.43 USD	–	–
40	Sonnet 4.5	high	3838.74 USD	–	–
41	Gemini 3.1 Pro	high	3774.25 USD	–	–
42	Gemini 3 Flash	high	3634.72 USD	–	–
43	Haiku 5.5	max	3602.46 USD	–	–
44	GPT-5.2	xhigh	3591.33 USD	–	–
45	DeepSeek V4 Pro	max	3284.52 USD	–	–
46	GLM 4.7	none	2376.82 USD	–	–
47	MiniMax M3	max	2157.77 USD	–	–
48	GPT-5.1	high	1473.43 USD	–	–
49	Kimi K2.5	max	1198.46 USD	–	–
50	Grok 4.1 Fast	high	1106.63 USD	–	–
51	DeepSeek V3.2	high	1034.00 USD	–	–
52	Gemini 2.5 Pro	high	573.64 USD	–	–
53	Gemini 2.5 Flash	medium	548.84 USD	–	–
54	Qwen 3.5 35B-A3B	low	462.69 USD	–	–
55	Haiku 4.5	xhigh	458.89 USD	–	–
56	Qwen 3.5 27B	high	201.98 USD	–	–
57	MiniMax M2	–	160.60 USD	–	–
58	Qwen 3 Max	max	71.57 USD	–	–
59	Grok 4.3	xhigh	35.26 USD	–	–
60	Qwen 3.5 Plus	none	0.54 USD	–	–
61	Qwen 3 235B-A22B	xhigh	-11.34 USD	–	–
62	gpt-oss-120b	high	-21.53 USD	–	–
63	MiniMax M2.5	high	-23.16 USD	–	–
64	GPT-5 Mini	high	-31.18 USD	–	–