FrontierCode 1.1 Extended · Open Analysis

#	Model	Effort	Score	Cost / task	Harness
1	Opus 5.5	medium	65.3%	$0.667	claude-code
2	Fable 5	xhigh	64.9%	$10.53	claude-code
3	GPT-6 Astra	max	64.5%	$3.93	codex
4	Sonnet 5.5	xhigh	64.4%	$1.24	claude-code
5	Opus 5	medium	63.6%	$3.51	claude-code
6	Fable 5.1	medium	63.6%	$2.68	claude-code
7	SWE 2	max	62.5%	$0.940	devin
8	Grok 4.6	high	61.3%	$2.38	grok-build
9	GPT-6 Sol	max	60.7%	$1.66	codex
10	Gemini 4	low	60.6%	$6.14	chisel
11	GPT-5.6 Sol	max	60.6%	$4.25	codex
12	GPT-6.1 Sol	medium	60.4%	$0.309	codex
13	Opus 4.8	max	59.6%	$8.05	claude-code
14	Grok 4.7	high	59.4%	$5.59	grok-build
15	Haiku 5.5	max	58.4%	$1.08	claude-code
16	Kimi K3	none	58.2%	$3.12	mini-swe-agent
17	GPT-5.5	xhigh	56.7%	$3.34	codex
18	DeepSeek V4.1 Flash	high	56.6%	$0.500	chisel
19	Grok 4.5	high	56.5%	$1.09	grok-build
20	Gemini 3.7 Flash	medium	56.3%	$1.53	chisel
21	Sonnet 5	xhigh	56.2%	$8.23	claude-code
22	GPT-6 Luna	max	56.1%	$0.083	codex
23	GPT-5.6 Terra	max	55.8%	$1.57	codex
24	GPT-5.6 Luna	max	55.1%	$0.307	codex
25	SWE 1.7	none	54.3%	$1.64	chisel
26	Opus 4.7	max	53.9%	$7.61	claude-code
27	Gemini 3.8 Flash	medium	53.4%	$2.21	chisel
28	Mistral Large 4	high	53.2%	$3.48	chisel
29	GLM 5.3	max	52.6%	$13.79	chisel
30	Gemini 3.6 Flash	medium	48.0%	$3.41	chisel
31	GLM 5.3 Flash	max	46.2%	$0.944	chisel
32	Kimi K2.7	none	45.4%	$2.56	mini-swe-agent
33	DeepSeek V4 Pro 0813	high	44.5%	$1.50	chisel
34	Opus 4.6	high	43.7%	$3.30	claude-code
35	GPT-5.4 Mini	xhigh	43.0%	$1.26	codex
36	Composer 2.5	none	40.8%	$2.57	cursor-cli
37	GLM 5.2	none	40.1%	$2.14	mini-swe-agent
38	Sonnet 4.6	max	40.0%	$2.44	claude-code
39	DeepSeek V4 Flash 0731	high	31.7%	$1.31	chisel
40	DeepSeek V4 Pro	none	31.0%	$1.37	mini-swe-agent
41	MiniMax M3	none	28.5%	$0.592	mini-swe-agent
42	Nemotron 3 Ultra 550B A55B	none	27.2%	$1.32	chisel
43	Inkling	max	24.8%	$3.20	mini-swe-agent
44	Qwen 3.7 Plus	none	21.8%	$0.212	mini-swe-agent
45	SWE 1.6	none	20.5%	$0.450	chisel
46	Mistral 3.5	none	18.7%	$1.24	chisel