OpenAI charges 2x for Fast and 6x for Ultrafast on its flagship models
OpenAI's API pricing page, read on 26 October 2026, lists gpt-6-astra at $10.00 input and $50.00 output per million tokens at standard, $20.00 and $100.00 at Fast, and $60.00 and $300.00 at Ultrafast. Batch is half of standard. The same page says GPT-5.6 Sol promotional pricing runs at least through 21 November 2026.
The price ladder: Standard, Batch, Fast, Ultrafast
OpenAI's pricing page lists the same flagship models at four speed and delivery tiers. The page read on 26 October 2026 lists three flagship models; the multiples in the tiles are computed by division from its tables.
Fast costs twice the standard input and output price on gpt-6-astra, gpt-6.1-sol and gpt-6-luna. OpenAI says priority processing was renamed Fast mode on 30 July 2026.
Ultrafast costs six times standard on gpt-6-astra and gpt-6.1-sol. gpt-6-luna has no Ultrafast row on the page.
Batch is half the standard price on all three flagship models.
OpenAI says regional processing (data residency) endpoints carry a 10% uplift for models released on or after 5 March 2026. FedRAMP endpoints carry one too.
OpenAI lists web search at this price per 1,000 calls for all models, plus search content tokens at model rates.
OpenAI says promotional pricing for GPT-5.6 Sol is available at least through 21 November 2026. The page does not state a price after that date.
| Per million tokens, input / output | Standard | Batch | Fast | Ultrafast |
|---|---|---|---|---|
| gpt-6-astra | $10.00 / $50.00 | $5.00 / $25.00 | $20.00 / $100.00 | $60.00 / $300.00 |
| gpt-6.1-sol | $2.00 / $10.00 | $1.00 / $5.00 | $4.00 / $20.00 | $12.00 / $60.00 |
| gpt-6-luna | $0.10 / $0.50 | $0.05 / $0.25 | $0.20 / $1.00 | not listed |
What a tier costs on a real job
The tier you pick changes the cost of the same job by up to twelve times. The table prices one example job on every listed tier.
| Example: 1 million input and 250,000 output tokens | Standard | Batch | Fast | Ultrafast |
|---|---|---|---|---|
| gpt-6-astra | $22.50 | $11.25 | $45.00 | $135.00 |
| gpt-6.1-sol | $4.50 | $2.25 | $9.00 | $27.00 |
| gpt-6-luna | $0.23 | $0.11 | $0.45 | not listed |
Assumption, not a measurement: 1 million input and 250,000 output tokens, uncached. Derived by us from the table above.
The GPT-5.6 Sol promotion
Below the flagship tables, OpenAI's page says promotional pricing for GPT-5.6 Sol is available at least through 21 November 2026. The static page does not say which listed row is the promotional one, and does not state what the price becomes. We therefore publish no price for after the promotion.
The same page lists a gpt-5.6-sol row under a separate Cyber models heading. We do not know whether that row is the promotional price.
How we know, and what we did not verify
Prices were parsed from a saved copy of developers.openai.com/api/docs/pricing (SHA-256 cf734fcb60d01a59), read on 26 October 2026. The build stops if the tables stop matching.
- We did not bill any usage or compare invoices. These are list prices as the page states them.
- Only the three flagship models and the web search tool are covered here. OpenAI lists many more models and tools.
- We have not asked OpenAI to comment before publication. A provider or reader who finds a wrong figure can tell us; we correct it and date the correction.
Part of the series: price calendar, search tool prices, DeepSeek pricing.
FAQ
What is the difference between Fast and Ultrafast on the OpenAI API?
On OpenAI's pricing page, read 26 October 2026, Fast is listed at 2x the standard price and Ultrafast at 6x for gpt-6-astra and gpt-6.1-sol. For gpt-6-astra that is $10.00 input and $50.00 output at standard, $20.00 and $100.00 at Fast, and $60.00 and $300.00 at Ultrafast, per million tokens. We did not measure the speed difference.
How much does Batch save?
Half. The page lists Batch at $1.00 input and $5.00 output for gpt-6.1-sol, against $2.00 and $10.00 at standard.
What happens to GPT-5.6 Sol pricing after 21 November?
OpenAI's page says promotional pricing is available at least through 21 November 2026 and does not state a later price, or which listed row is the promotional one. We will not guess. The page is re-read weekly.
Does OpenAI charge more for regional processing?
Yes, per its page: regional processing (data residency) endpoints are charged a 10% uplift for models released on or after 5 March 2026 that are eligible for data residency, and FedRAMP endpoints are also charged a 10% uplift.
What does OpenAI charge for web search?
The page lists web search for all models at $10.00 per 1,000 calls plus search content tokens billed at model rates. See our comparison of search tool prices across providers.