AnalysisAI ModelsAugust 18, 2026

DeepSeek V4 Pro 0813 vs GPT-5.6 Sol on DeepSWE: Cost, Coding, and Routing

On 904 DeepSWE rollouts, GPT-5.6 Sol leads pass@1 (72.7% vs 62.8%) but costs 35x more ($8.37 vs $0.24 per rollout); DeepSeek V4 Pro 0813 wins pass@4 (88.5% vs 85.8%). A Pro-first cascade hits 83.0% of tasks at $3.35 each.

1 source

Daily brief

Get tomorrow's AI brief in your inbox

More stories today

Open the live feed
DeepSeek V4 Pro 0813 vs GPT-5.6 Sol on DeepSWE: Cost, Coding, and Routing — AIBriefs