OPENAI · MODEL TELEMETRY CARD
GPT-5.2
Canonical UID: openai/gpt-5.2-20251211 · Released: 2025-12-11
| Domain | Benchmark | Effort / Mode | ⚡ Value/Cost # | 🏆 Performance # | Score | Task $ | Unit Cost $ | Unit |
|---|---|---|---|---|---|---|---|---|
| Code | ALE-Bench | MEDIUM | #61 / 94 | #33 / 94 | 1249.83 | $0.142947 | $0.0001144 | rating point |
| Code | ALE-Bench | HIGH | #76 / 94 | #31 / 94 | 1293.55 | $0.296865 | $0.0002295 | rating point |
| Reason | ProofBench | XHIGH | #59 / 62 | #51 / 62 | 0.15 | $1.05969 | $7.065 | correct answer |
Compare against all models on ← #frontier Live Ranking Board · Fixed-capability cost half-life analysis on /methodology.