OpenAI’s newest frontier model for the hardest coding, agent, and alignment-sensitive work. Use when GPT-5.6 Sol is not enough and you can afford Astra list rates plus cache and Fast mode multipliers.
Privacy first counting
TokenCalculator counts tokens in your browser when an Exact encoding is available. Prompt text does not need to leave your device for local counting.
Pricing snapshot
USD per 1 million tokens. Verify against the provider source before you lock budget.
| Rate | USD |
|---|---|
| Input / 1M | $10.00 |
| Output / 1M | $50.00 |
| Cached input / 1M | $1.00 |
| Cache write / 1M | $12.50 |
| Batch input / 1M | $5.00 |
| Batch output / 1M | $25.00 |
| Long-context input (≥ 272K prompt) | $20.00 |
| Long-context output (≥ 272K prompt) | $75.00 |
| Long-context cached input (≥ 272K prompt) | $2.00 |
Model details
| Detail | Value |
|---|---|
| Provider | OpenAI |
| Context window | 1.1M |
| Max output | 128K |
| Tokenizer | Exact (o200k_base) |
| Last verified | 2026-09-18 |
| API model id | gpt-6-astra |
Notes
Standard rates for prompts under 272K input tokens. Longer prompts reprice the entire request at the long-context band. Fast mode is 2× standard. Cache writes are billed separately from cache reads.
More OpenAI models
Same provider catalog. Use these when you are comparing rates within one stack.
- GPT-4 Turbo: $10.00 in / $30.00 out
- ChatGPT-4o Latest: $5.00 in / $15.00 out
- o1: $15.00 in / $60.00 out
- o1-preview: $15.00 in / $60.00 out
- GPT-5.6 Sol: $4.00 in / $20.00 out
- GPT-4o: $2.50 in / $10.00 out
- GPT-4o (2024-08-06): $2.50 in / $10.00 out
- GPT-4.1: $2.00 in / $8.00 out
Featured comparisons
Browse the full set on Compare.
Related tools
- Estimate API cost
- Count tokens
- Check context fit
- Estimate cache savings
- Compare batch pricing
- Token visualizer