S
Sync
Dev
Home
About
Pricing
Blogs
Contact
Start a project
Home
About
Pricing
Blogs
Contact
Start a project
Home
/
Benchmarks
/
SWE-Bench Verified
SWE-Bench Verified
Software Engineering Benchmark (Verified)
Model ranking
#
Model
Score (%)
1
Claude 3.5 Sonnet
Anthropic
49.0%
2
Gemini 1.5 Pro
Google DeepMind
38.0%
3
GPT-4o
OpenAI
33.2%
4
Llama 3.1 405B
Meta AI
24.0%