GPT-6 Astra
GPT-6 Astra has 5 tested configurations. Its highest published overall score is 94.7 at xhigh. Inspect dimension scores, evaluation time, and cost below.
Overall score94.7
GPT-6 Astra · max
- Backend engineering
- 95
- Frontend interaction
- 94
- Knowledge & reasoning
- 95
- Evaluation time
- 93m 26s
- Cost
- $8.35
Complete evaluation · API cost
Results by reasoning effort
Swipe to see all results →
| Model / effort | Overall | Backend engineering | Frontend interaction | Knowledge & reasoning | Total eval. time | Total ref. cost | Compare |
|---|---|---|---|---|---|---|---|
| 88.3 | 88 | 92 | 85 | 61m 27s | $3.70 | ||
| 88.0 | 91 | 87 | 85 | 72m 41s | $3.76 | ||
| 91.1 | 92 | 96 | 85 | 79m 52s | $4.40 | ||
| 94.7 | 95 | 94 | 95 | 83m 56s | $7.01 | ||
| 94.7 | 95 | 94 | 95 | 93m 26s | $8.35 |
Score breakdown
- Overall score
- 94.7
- Evaluation time
- 93m 26s
- Cost
- $8.35
Service logic, error handling and verification.
Contracts and boundaries19 / 20
API contracts, input validation and boundary behavior.
Retries and failures20 / 20
Recovery, retry conditions and failure handling.
CI and adversarial review19 / 20
Verification plans, risk detection and audit evidence.
Transactions and consistency20 / 20
State transitions, concurrency and consistency.
Caching and regression17 / 20
Invalidation, data updates and regression coverage.
What changes with higher reasoning effort
Compare overall scores, cost, and time across neighboring tested efforts.
xhigh max
- Overall
- 0.0 pts 94.7 → 94.7
- Cost
- +$1.34 $7.01 → $8.35
- Evaluation time
- +11.3% 83m 56s → 93m 26s
Data snapshot published · 11/13/2026Scoring & cost notes