Harness
codex
codex is strongest with qwen3.8-flash, 4th overall.
Best overall rank4th
Best partner
qwen3.8-flash
qwen3.8-flashStrongest inPresentations
First places10 tasks
Pairings evaluated11 of 19 models
Time and cost, best pairingTime1505.3sCost$0.1449
Who codex works best with
Valid evaluations covered 11 of 19 models. The other 8 have not been evaluated yet.
- qwen3.8-flash1096#4
glm-5.3-prime1077#7- qwen3.8-max1065#8
gpt-5.6-terra994#20- gpt-6-sol994#21
- glm-5.3992#22
- gpt-6.1-sol974#26
- gpt-6-astra974#27
- qwen3.7-plus971#30
- gpt-5.6-sol918#35
claude-opus-5-5—Not evaluated
kimi-k3—Not evaluated- claude-fable-5-1—Not evaluated
deepseek-flash—Not evaluated- claude-sonnet-5—Not evaluated
- kimi-k2.7-code—Not evaluated
- deepseek-v4-pro—Not evaluated
- claude-haiku-4-5-20251001—Not evaluated
Places across 24 tasks
- 101st
- 42nd
- 43rd
- 04th
- 25th
- 46th and below
Most often 1st, in 10 tasks.
Rank in each kind of work
Each row is a kind of work and each dot is one pairing's place in it; further left is better. Hover a dot for its name and place, or pick a name below to follow one pairing.
Where it places in each task
The number in each cell is a place: where that pairing ranked in that task, 1 being best, so 3 means 3rd. Hover a number to see how many pairings were ranked there. A dash means the pairing was not run on that task.
| Task | qwen3.8-flash | glm-5.3-prime | qwen3.8-max | gpt-5.6-terra | gpt-6-sol | glm-5.3 | gpt-6.1-sol | gpt-6-astra | qwen3.7-plus | gpt-5.6-sol |  MiniMax-M2.5 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Academic research | 15 | 11 | 14 | 22 | 16 | 33 | 12 | 1 | 29 | 20 | 32 |
| IT management | 15 | 12 | 1 | 15 | 27 | 1 | 15 | 1 | 1 | 15 | 35 |
1st2nd3rd4th5th6th and below