Clef
Cloudflare · cloudflare/clef · added 20 Oct 2026
Clef by Cloudflare ranks #6 of 12 AI decision models overall on DecisionBench: #6 of 12 on Sudoku, #8 of 12 on Triage, #4 of 12 on Robustness. View on OpenRouter.
- Overall
- #6 of 12
- Average rank
- 6.0
- Sudoku
- #6 of 12
- Triage
- #8 of 12
- Robustness
- #4 of 12
Sudoku Logical reasoning
Leaderboard →- Rank
- #6 of 12
- Accuracy
- 80.0%
- Solved
- 20/20
- Mistakes / puzzle
- 12.6
- Solve time / puzzle
- 23s
- Cost / puzzle
- $0.0101
By difficulty
| Difficulty | Rank | Accuracy | Solved | Mistakes | Solve time | Cost |
|---|---|---|---|---|---|---|
| Easy | #5 | 90.3% | 5/5 | 4.4 | 11s | $0.004689 |
| Medium | #6 | 82.5% | 5/5 | 10.4 | 32s | $0.0139 |
| Hard | #3 | 79.6% | 5/5 | 13.8 | 22s | $0.0103 |
| Expert | #8 | 72.6% | 5/5 | 21.6 | 29s | $0.0113 |
Every race (20)
| Puzzle | Mistakes | Calls | Solve time | Cost | Status |
|---|---|---|---|---|---|
| easy-01 | 2 | 15 | 9.6s | $0.005031 | Solved |
| easy-02 | 9 | 19 | 13s | $0.006169 | Solved |
| easy-03 | 4 | 16 | 11s | $0.005173 | Solved |
| easy-04 | 3 | 16 | 10s | $0.003544 | Solved |
| easy-05 | 4 | 16 | 10s | $0.003528 | Solved |
| medium-01 | 15 | 28 | 21s | $0.008855 | Solved |
| medium-02 | 10 | 24 | 18s | $0.007552 | Solved |
| medium-03 | 5 | 19 | 37s | $0.0149 | Solved |
| medium-04 | 18 | 30 | 59s | $0.0295 | Solved |
| medium-05 | 4 | 20 | 25s | $0.008838 | Solved |
| hard-01 | 17 | 31 | 24s | $0.009269 | Solved |
| hard-02 | 9 | 25 | 20s | $0.009525 | Solved |
| hard-03 | 9 | 23 | 19s | $0.0108 | Solved |
| hard-04 | 17 | 31 | 24s | $0.0105 | Solved |
| hard-05 | 17 | 31 | 23s | $0.0115 | Solved |
| expert-01 | 31 | 43 | 33s | $0.0151 | Solved |
| expert-02 | 19 | 33 | 26s | $0.0103 | Solved |
| expert-03 | 24 | 38 | 32s | $0.0131 | Solved |
| expert-04 | 21 | 35 | 30s | $0.009962 | Solved |
| expert-05 | 13 | 28 | 22s | $0.007962 | Solved |
Triage Following written rules
Leaderboard →- Rank
- #8 of 12
- Accuracy
- 87.2%
- Team
- 100.0%
- Refund
- 96.7%
- Urgency
- 65.0%
- Confidence gap
- 0.057
60 tickets · 4.3% of mistakes made while at least 90% sure · $0.0451 per 1,000 tickets
Robustness Consistency under pressure
Leaderboard →- Rank
- #4 of 12
- Score
- 99.5%
- Same if re-asked
- 99.2%
- Changed by reorder
- 0.0%
- In a list of 40
- 87% / 85%
- Obeyed planted text
- 1.7%
20 tickets × 6 checks · changed by rename 0.0% · accuracy with planted text 85.0%