Use PyGhidra Python bindings for binary decompilation.
Performance
| Model | Pass Rate | Runs | Avg Cost | Avg Time |
|---|---|---|---|---|
| glm-4.7 | 100% | | $0.08 | 12m |
| glm-5 | 100% | | $0.09 | 11m |
| claude-opus-4.5 | 100% | | $0.14 | 3m |
| gpt-5.3-codex | 100% | | $0.15 | 3m |
| gemini-3.1-pro-preview | 100% | | $0.18 | 4m |
| claude-opus-4.6 | 100% | | $0.19 | 4m |
| gemini-3-pro-preview | 100% | | $0.19 | 4m |
| gpt-5.3-codex-xhigh | 100% | | $0.22 | 5m |
| gpt-5.3-codex-high | 100% | | $0.24 | 6m |
| gpt-5 | 100% | | $0.25 | 11m |
| gpt-5.2-codex-high | 100% | | $0.26 | 6m |
| claude-sonnet-4.5 | 100% | | $0.27 | 7m |
| gpt-5.2-codex-xhigh | 100% | | $0.28 | 6m |
| gpt-5.2 | 100% | | $0.28 | 8m |
| kimi-k2.5 | 67% | | $0.02 | 4m |
| deepseek-v3.2 | 67% | | $0.02 | 8m |
| gpt-5.2-codex | 67% | | $0.30 | 8m |
| grok-4 | 67% | | $0.32 | 11m |
| claude-sonnet-4 | 67% | | $0.32 | 6m |
| gemini-3-flash-preview | 33% | | $0.02 | 2m |
| grok-4.1-fast | 33% | | $0.05 | 24m |
| gemini-2.5-pro | 33% | | $0.31 | 8m |
| claude-haiku-4.5 | 33% | | $0.38 | 15m |
All product names, logos, and brands (™/®) are the property of their respective owners; they're used here solely for identification and comparison, and their use does not imply affiliation, endorsement, or sponsorship.