Augment Code

🔓 Open

Enterprise-grade agentic coding platform with deep codebase understanding, multi-repo context, and persistent workspace memory.

💰 Open-source · vscode, jetbrains, web

70.0Overall Score

Non-Gameable Scoring

Scores are derived from established benchmarks, adjusted for harness-specific performance across four dimensions: Coding, Reasoning, Tool Use, and Autonomy.

Each dimension starts from public benchmark data and applies harness-specific modifiers based on tool integration, context handling, and orchestration quality. The overall score is a weighted composite that penalizes narrow optimization.

Model	Overall	Coding	Reasoning	Tool Use	Autonomy
Claude Opus 4.6	70.0	75.38	87.25	52.15	68.35
Gemini 3.1 Pro	59.4	81.19	82.49	44.93	56.17
Kimi K2.5	59.1	75.61	75.27	30.19	56.74
GPT-5.4	52.8	47.9	64.69	64.54	60.84
Claude Sonnet 4.6	52.7	69.37	54.65	48.62	65.57
MiniMax-M2.7	51.0	62.68	50.51	49.25	59.09
DeepSeek R1	46.4	71.2	50.51	32.94	41.22
Qwen 3.5	45.1	53.05	50.97	22.44	42.36
Gemini 3 Pro	40.9	44.39	54.94	35.31	43.38
MiMo-V2-Flash	40.1	65.01	46.74	27.71	35.11