DECK-Bench (Internal)Browse 284 families
DECK-Bench (Internal)
Can the model produce useful work in a professional knowledge-work setting?
Version / measure
- Domain
- Knowledge Work
- Ranked models
- 6
- Reported values
- 6
- Reports
- 1 · lab
- Direction
- Higher is better
- Status
- Version not specified
Rankings
Best reported score per model across effort settings
| # | Model | Best setting | Score |
|---|---|---|---|
| 1 | GPT-5.6 SolOpenAI | maxcost n/a1 value · Jul 17 · source | 74.7% |
| 2 | Kimi K3Moonshot AIOW | maxcost n/a1 value · Jul 17 · source | 73.5% |
| 3 | Claude Fable 5Anthropic | maxcost n/a1 value · Jul 17 · source | 73.0% |
| 4 | GLM-5.2Z.aiOW | maxcost n/a1 value · Jul 17 · source | 68.6% |
| 5 | GPT-5.5OpenAI | xhighcost n/a1 value · Jul 17 · source | 68.2% |
| 6 | Claude Opus 4.8Anthropic | maxcost n/a1 value · Jul 17 · source | 66.9% |
Effort curve
Score vs. reported cost. Lines connect complete sweeps only.
No cost-linked effort sweep for this version.
Definition and comparison boundaryinternal methodology
Can the model produce useful work in a professional knowledge-work setting?
Professional analysis, document, finance, office, or domain-specific tasks. Higher is better. The value is the percentage reported in this lab's table.
This is a publisher-defined internal evaluation. The task set or grading details are not fully public, so treat it as directional evidence.
Method / source ↗