DeepSWE

Measuring frontier coding agents on original, long-horizon engineering tasks

Get notified when new models drop

Leaderboard

113 tasks · updated July 1, 2026
DeepSWE score$0$5.00$10$15$20$250%10%20%30%40%50%60%70%80%Avg cost per taskmost efficient ↗claude-fable-5 [high]Defaultgpt-5.5 [medium]Defaultclaude-sonnet-5 [max]Defaultclaude-opus-4.8 [high]Defaultgpt-5.4 [xhigh]glm-5.2 [max]Defaultgemini-3.5-flash [medium]kimi-k2.7-codeclaude-sonnet-4.6 [high]gemini-3.1-pro [high]
claude-fable-5[max]
70%±4%
Avg cost $21.63Out tok 119kSteps 88
gpt-5.5[xhigh]
67%±6%
Avg cost $7.23Out tok 46kSteps 82
claude-opus-4.8[max]
59%±2%
Avg cost $13.22Out tok 135kSteps 120
claude-sonnet-5[max]
54%±4%
Avg cost $26.40Out tok 214kSteps 268
gpt-5.4[xhigh]
52%±2%
Avg cost $5.65Out tok 71kSteps 70
glm-5.2[max]
44%±2%
Avg cost $3.92Out tok 78kSteps 129
gemini-3.5-flash[medium]
37%±2%
Avg cost $7.34Out tok 276kSteps 86
kimi-k2.7-code
31%±1%
Avg cost $2.82Out tok 59kSteps 149
claude-sonnet-4.6[high]
30%±4%
Avg cost $5.52Out tok 76kSteps 134
gemini-3.1-pro[high]
12%±2%
Avg cost $9.48Out tok 196kSteps 81
0%20%40%60%80%

All models run on mini-swe-agent for consistency. Read why.

Today's leading public coding benchmarks are starting to saturate at the frontier: top models cluster within a narrow score band where adjacent configurations often overlap on confidence intervals. DeepSWE is a long-horizon software engineering benchmark built to separate them. It delivers four advances over existing public benchmarks:

  • Contamination free: Tasks are written from scratch, not adapted from existing commits or PRs, so no model has seen the solution during pretraining.
  • High diversity: Tasks span a broad pool of 91 repositories across 5 languages.
  • Real-world complexity: Prompts are ~half the length of SWE-bench Pro's, yet solutions require 5.5x more code and ~2x more output tokens.
  • Reliable verification: Verifiers are hand-written to test software behavior rather than implementation details.

The result is a benchmark that reflects how today's frontier coding agents actually perform in software engineering work.

Task Examples

All 113 tasks

Sign up for leaderboard updates

New frontier models are added to the DeepSWE leaderboard as they're released.