All billing dimensions
Live pricing is fetched from the OneInfer catalog where available; the cache-write rate and the long-context surcharge are not present in the live pricing record and are stated here as sourced, dated facts instead.
| Dimension | Rate | Source |
|---|---|---|
| Input / 1M tokens | $10.00 | Live catalog |
| Output / 1M tokens | $50.00 | Live catalog |
| Cached input / 1M tokens | $1.00 | Live catalog |
| Cache write / 1M tokens | $12.50 | OpenAI Platform pricing docs (not in live pricing record) |
The long-context surcharge, above 272,000 input tokens
Above a 272,000-token input threshold, GPT-6: Astra prices at 2x the standard input rate and 1.5x the standard output rate. This is a published multiplier rather than a separate live-fetchable tier: at $10 and $50 standard, the surcharged rates work out to $20 per 1M input and $75 per 1M output.
| Tier | Input $/1M | Output $/1M |
|---|---|---|
| Up to 272,000 input tokens | $10.00 | $50.00 |
| Above 272,000 input tokens | $20.00 (2x) | $75.00 (1.5x) |
Fast Mode
Fast Mode runs at 2x the applicable price (standard or long-context) for up to 2x the speed. It is a latency lever, not a quality change — reserve it for workloads where response time is the binding constraint.
Ready to test the workflow?
Create account & add creditsBatch and caching
GPT-6: Astra supports batch requests and prompt caching. Cached input at $1.00 per 1M tokens is ten times the price of Claude Fable 5.1's $0.25 cache-read rate at an identical list price elsewhere on the ledger — a cache-heavy agent will cost meaningfully more on Astra than on a competitor with a lower cache-read rate.
Frequently asked questions
How much does GPT-6: Astra cost?
GPT-6: Astra costs $10 per 1M input tokens and $50 per 1M output tokens on OneInfer. Cached input reads at $1.00 per 1M tokens, and cache writes cost $12.50 per 1M. Above 272,000 input tokens, a 2x input / 1.5x output surcharge applies. Fast Mode runs at 2x the applicable price for up to 2x the speed.
Is there a long-context surcharge on GPT-6: Astra?
Yes. Above 272,000 input tokens, GPT-6: Astra prices at 2x the standard input rate ($20 per 1M) and 1.5x the standard output rate ($75 per 1M). This is a published multiplier, not a separately live-fetchable pricing tier.
Put GPT-6: Astra to work
Fund a controlled evaluation, start with a prepared prompt, and measure quality and cost on your own workload.