Leaderboard

Every kind of work is ranked on its own. See which harness and model deliver best together.

Documents

Document processing

Pairings
49
Tasks
11
Order
Highest quality first

Quality and cost

Top left is best. On the line: no other pairing is both cheaper and better.

↑ Quality
600800100012001400

On the line

  1. opencode × deepseek-flashQuality999Cost$0.0036
  2. codex × deepseek-flashQuality1013Cost$0.0080
  3. deepseek_harness × deepseek-flashQuality1049Cost$0.0093
  4. opencode × glm-5.3Quality1085Cost$0.0191
  5. opencode × claude-sonnet-5Quality1164Cost$0.0453
  6. codex × gpt-5.6-terraQuality1189Cost$0.0534
  7. codex × gpt-6.1-solQuality1265Cost$0.0599
  8. opencode × gpt-6.1-solQuality1277Cost$0.0933

All pairings

  1. 1opencode×gpt-6.1-solQuality 1277Time 109.6sCost $0.0933
  2. 2codex×gpt-6.1-solQuality 1265Time 114.3sCost $0.0599
  3. 3opencode×claude-fable-5-1Quality 1257Time 99.7sCost $0.1511
  4. 4codex×gpt-6-astraQuality 1250Time 277.4sCost $0.1829
  5. 5opencode×gpt-6-astraQuality 1250Time 189sCost $0.2525
  6. 6claude_code×claude-fable-5-1Quality 1219Time 70.7sCost $0.1111
  7. 7codex×gpt-6-solQuality 1212Time 93.7sCost $0.0654
  8. 8opencode×gpt-5.6-solQuality 1209Time 120.4sCost $0.2092
  9. 9opencode×gpt-6-solQuality 1206Time 163.1sCost $0.1175
  10. 10claude_code×claude-opus-5-5Quality 1205Time 79.3sCost $0.2216
  11. 11codex×gpt-5.6-solQuality 1205Time 112.7sCost $0.1279
  12. 12opencode×gpt-5.6-terraQuality 1194Time 124.9sCost $0.0844
  13. 13opencode×claude-opus-5-5Quality 1194Time 99.3sCost $0.1504
  14. 14codex×gpt-5.6-terraQuality 1189Time 80.1sCost $0.0534
  15. 15opencode×claude-sonnet-5Quality 1164Time 105.7sCost $0.0453
  16. 16opencode×glm-5.3Quality 1085Time 88.3sCost $0.0191
  17. 17codex×kimi-k3Quality 1069Time 145sCost $0.0432
  18. 18opencode×kimi-k3Quality 1053Time 98.1sCost $0.0396
  19. 19deepseek_harness×deepseek-flashQuality 1049Time 69.6sCost $0.0093
  20. 20kimi_cli×kimi-k3Quality 1030Time 193sCost $0.1818