Decider Flash v1.2
Perplexity · perplexity/pplx-decider-flash-v1.2 · added 20 Oct 2026
Decider Flash v1.2 by Perplexity ranks #11 of 12 AI decision models overall on DecisionBench: #12 of 12 on Sudoku, #5 of 12 on Triage, #9 of 12 on Robustness. View on OpenRouter.
- Overall
- #11 of 12
- Average rank
- 8.7
- Sudoku
- #12 of 12
- Triage
- #5 of 12
- Robustness
- #9 of 12
Sudoku Logical reasoning
Leaderboard →- Rank
- #12 of 12
- Accuracy
- 79.1%
- Solved
- 16/20
- Mistakes / puzzle
- 11.2
- Solve time / puzzle
- 37s
- Cost / puzzle
- $0.004203
By difficulty
| Difficulty | Rank | Accuracy | Solved | Mistakes | Solve time | Cost |
|---|---|---|---|---|---|---|
| Easy | #7 | 89.5% | 5/5 | 4.8 | 22s | $0.002428 |
| Medium | #8 | 81.4% | 5/5 | 11.2 | 33s | $0.004115 |
| Hard | #11 | 78.4% | 4/5 | 12.2 | 37s | $0.004642 |
| Expert | #12 | 68.3% | 2/5 | 16.6 | 56s | $0.005625 |
Every race (20)
| Puzzle | Mistakes | Calls | Solve time | Cost | Status |
|---|---|---|---|---|---|
| easy-01 | 5 | 16 | 22s | $0.002362 | Solved |
| easy-02 | 3 | 16 | 21s | $0.002229 | Solved |
| easy-03 | 5 | 18 | 24s | $0.002602 | Solved |
| easy-04 | 8 | 18 | 25s | $0.002809 | Solved |
| easy-05 | 3 | 16 | 20s | $0.002139 | Solved |
| medium-01 | 16 | 25 | 35s | $0.004489 | Solved |
| medium-02 | 11 | 24 | 33s | $0.004072 | Solved |
| medium-03 | 9 | 22 | 30s | $0.003641 | Solved |
| medium-04 | 9 | 24 | 32s | $0.003893 | Solved |
| medium-05 | 11 | 24 | 35s | $0.004481 | Solved |
| hard-01 | 9 | 24 | 35s | $0.004222 | Solved |
| hard-02 | 14 | 29 | 43s | $0.005290 | Solved |
| hard-03 | 11 | 26 | 40s | $0.004848 | Solved |
| hard-04 | 6 | 6 | 12s | $0.001409 | Error |
| hard-05 | 21 | 34 | 55s | $0.007441 | Solved |
| expert-01 | 26 | 40 | 64s | $0.008781 | Solved |
| expert-02 | 19 | 33 | 55s | $0.006908 | Solved |
| expert-03 | 20 | 31 | 80s | $0.007381 | Error |
| expert-04 | 4 | 5 | 24s | $0.001138 | Error |
| expert-05 | 14 | 16 | 55s | $0.003916 | Error |
Triage Following written rules
Leaderboard →- Rank
- #5 of 12
- Accuracy
- 97.2%
- Team
- 100.0%
- Refund
- 100.0%
- Urgency
- 91.7%
- Confidence gap
- 0.016
60 tickets · 60.0% of mistakes made while at least 90% sure · $0.0168 per 1,000 tickets
Robustness Consistency under pressure
Leaderboard →- Rank
- #9 of 12
- Score
- 97.3%
- Same if re-asked
- 100.0%
- Changed by reorder
- 0.0%
- In a list of 40
- 95% / 98%
- Obeyed planted text
- 10.0%
20 tickets × 6 checks · changed by rename 0.0% · accuracy with planted text 70.0%