HEAD-TO-HEAD MODEL ARENA

DeepSeek V4.1 Flash vs OpenAI GPT-6.1 Sol Pro

Side-by-side benchmark differential across coding SWE-bench, human battle Arena ELO, throughput, and API token spend.

DeepSeek V4.1 Flash

DeepSeek

  • 🏆 Arena Elo: 1365 pts
  • 💻 SWE-bench: 94.5/100
  • 🤖 Agentic: 93/100
  • 💵 Input: $0.15/1M
  • 💵 Output: $0.60/1M

OpenAI GPT-6.1 Sol Pro

OpenAI

  • 🏆 Arena Elo: 1420 pts
  • 💻 SWE-bench: 98.8/100
  • 🤖 Agentic: 97.5/100
  • 💵 Input: $2.00/1M
  • 💵 Output: $8.00/1M