Model pricing
GPT-6.2 Astra pricing
GPT-6.2 Astra by OpenAI costs $8 per 1M input tokens and $40 per 1M output tokens (cached input $0.4, batch input $4) as of 2026-11-22. A 1,000-input/500-output request costs $0.028. The lowest-cost alternative with at least this context window and the same recorded reasoning category is GPT-6 Luna Pro at $0.1/$0.5; capability equivalence is not guaranteed.
Notes
Short context ≤272K; long context >272K. Batch 50% off Standard; Fast 2× Standard. Ultrafast is 3× Standard through 28 Feb 2027 Pacific time and 6× from 1 Mar 2027. Background requests are 30% off for 2-hour and 40% off for 8-hour completion. No Flex tier.
Key pricing facts
| Fact | Value |
|---|---|
| Input / 1M tokens | $8 |
| Output / 1M tokens | $40 |
| Cached input / 1M | $0.4 |
| Batch input / output / 1M | $4 / $20 |
| Context window | 1,050,000 |
| Provider | OpenAI |
| Last verified | 2026-11-21 |
| Snapshot updated | 2026-11-21 |
| Long prompt >272,000 tokens / 1M | $16 input / $0.8 cached / $60 output |
| Long-prompt Batch input / output / 1M | $8 / $30 |
| 30-minute cache write (short / long) / 1M | $10 / $20 |
| 12-hour cache write (short / long) / 1M | $20 / $40 |
| Fast mode input / output / 1M | $16 / $80 |
| Ultrafast input / output / 1M | $24 / $120 |
| Background discounts | 30% off (2-hour) / 40% off (8-hour) |
Last updated: · Source: OpenAI pricing · Methodology
Frequently asked questions
What does GPT-6.2 Astra cost per million tokens?
Input costs $8 and output costs $40 per million tokens as of 2026-11-22T01:13:10.749Z. The 1,000-input/500-output example costs $0.028, before taxes, hidden reasoning tokens or unselected service-tier surcharges.
Does GPT-6.2 Astra have cached or batch pricing?
Cached input: $0.4 per million tokens under the recorded cache rules. Batch: $4 input and $20 output per million tokens. Eligibility and cache-write charges depend on the provider; these discounts are not automatically stacked.
What is a cheaper alternative to GPT-6.2 Astra?
GPT-6 Luna Pro costs $0.1 input and $0.5 output per million tokens. This comparison matches context capacity and reasoning category, not benchmark quality. Test your workload before replacing a model.
How often are these prices verified?
Curated catalog counts match the pricing table and public snapshot, not a live database census. Dates change only when recorded pricing or model facts change; a daily synchronization is not independent provider verification. Snapshot updated 2026-11-21. The recorded model verification date is 2026-11-21.