SShogo Bench

FrontierCode

Frontier-difficulty programming problems.

Agent Index weight
0%
shown for context only
Models tested
46
from the benchmark’s leaderboard and reports
Normalised for the index
0 → 0, 1 → 100
Official site

Results

#ModelScoreNormalisedHarnessCost/taskDate
1Claude Opus 5.5 Supported
Anthropic
54.6%claude-code——
2Claude Fable 5 Supported
Anthropic
53.5%claude-code——
3Claude Opus 5 Supported
Anthropic
53.4%claude-code——
4GPT-6 Astra Supported
OpenAI
53.3%codex——
5Claude Sonnet 5.5 Estimated
Anthropic
52.1%claude-code——
6Claude Fable 5.1 Supported
Anthropic
50.9%claude-code——
7GPT-6.1 Sol Estimated
OpenAI
50.2%codex——
8SWE-2
Cognition
50.0%devin——
9GPT-6 Sol Estimated
OpenAI
49.3%codex——
10Grok 4.6 Supported
xAI
48.0%grok-build——
11Grok 4.7 Estimated
xAI
47.6%grok-build——
12Gemini 4 Argon Supported
Google
47.5%devin-cli——
13GPT-5.6 Sol Supported
OpenAI
47.5%codex——
14Claude Opus 4.8 Supported
Anthropic
46.5%claude-code——
15Claude Haiku 5.5 Estimated
Anthropic
46.4%claude-code——
16Kimi K3 Supported
Moonshot AI
44.2%mini-swe-agent——
17DeepSeek V4.1 Flash Estimated
DeepSeek
43.9%devin-cli——
18Gemini 3.7 Flash Estimated
Google
43.6%devin-cli——
19GPT-5.5 Supported
OpenAI
43.0%codex——
20Claude Sonnet 5 Supported
Anthropic
42.7%claude-code——
21Grok 4.5 Estimated
xAI
42.4%grok-build——
22GPT-6 Luna
OpenAI
42.4%codex——
23SWE-1.7
Cognition
42.0%devin-cli——
24GPT-5.6 Terra Estimated
OpenAI
41.3%codex——
25Gemini 3.8 Flash Supported
Google
41.2%devin-cli——
26GLM-5.3 Supported
Z.ai (Zhipu)
40.1%devin-cli——
27GPT-5.6 Luna Estimated
OpenAI
39.8%codex——
28Mistral Large 4
Mistral
39.4%chisel——
29Claude Opus 4.7 Supported
Anthropic
38.5%claude-code——
30Gemini 3.6 Flash Estimated
Google
34.4%devin-cli——
31GLM-5.3-Flash Supported
Z.ai (Zhipu)
31.8%devin-cli——
32Kimi K2.7 Code Estimated
Moonshot AI
30.1%mini-swe-agent——
33DeepSeek V4 Pro 0813
DeepSeek
28.5%devin-cli——
34GPT-5.4 mini
OpenAI
27.0%codex——
35Claude Opus 4.6 Supported
Anthropic
26.6%claude-code——
36Composer 2.5
Cursor
25.6%cursor-cli——
37GLM-5.2 Supported
Z.ai (Zhipu)
24.5%mini-swe-agent——
38Claude Sonnet 4.6 Supported
Anthropic
24.3%claude-code——
39DeepSeek V4 Flash 0731
DeepSeek
18.8%devin-cli——
40DeepSeek V4 Pro Estimated
DeepSeek
17.6%mini-swe-agent——
41MiniMax-M3 Estimated
MiniMax
14.7%mini-swe-agent——
42Inkling Estimated
thinkingmachines
14.0%mini-swe-agent——
43Nemotron 3 Ultra
NVIDIA
13.6%devin-cli——
44Qwen3.7 Plus
Alibaba (Qwen)
10.2%mini-swe-agent——
45SWE-1.6
Cognition
9.4%devin-cli——
46Mistral Medium 3.5
Mistral
8.0%devin-cli——