· CostMyTokens · Analysis · 3 min read
GPT-6 API Pricing 2026: Astra vs GPT-6.1 Sol vs Luna
GPT-6 API pricing for Astra, GPT-6.1 Sol and Luna from OpenAI's official price list, with cached, batch, long-context and monthly costs worked out.
OpenAI’s current GPT-6 family has three tiers: GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna. This page lists their official Standard-tier prices per 1M tokens, with source and verification date, and works out what typical monthly workloads cost. Every number comes from the same data as our calculator.
Disclosure: this article contains no affiliate links. Where CostMyTokens does use affiliate links elsewhere, they never affect order — every table is sorted by price only. Read our disclosure.
GPT-6 API prices (official, per 1M tokens)
Standard tier and short context (≤272K input tokens), sorted by input price.
| Model | Input / 1M | Output / 1M | Cached input / 1M | Batch input / output | Context | Source |
|---|---|---|---|---|---|---|
| GPT-6 Luna | $0.1 | $0.5 | $0.01 | $0.05 / $0.25 | 1,050,000 | official pricing, verified 2026-10-25 |
| GPT-6.1 Sol | $2 | $10 | $0.1 | $1 / $5 | 1,050,000 | official pricing, verified 2026-10-25 |
| GPT-6 Astra | $10 | $50 | $1 | $5 / $25 | 1,050,000 | official pricing, verified 2026-10-25 |
Batch rates come from OpenAI’s Flex/Batch price table, which lists the same rates for both. Cache writes are billed separately: 1.25x input for a 30-minute write and 2.5x input for a 12-hour write.
What a month costs
Each workload is 80% input and 20% output, at the Standard tier. Sorted by cost.
| Model | 10M tokens / month | 100M tokens / month | Same, 50% of input cached | Same, via Batch API |
|---|---|---|---|---|
| GPT-6 Luna | $1.80 | $18.00 | $1.44 (10M) | $0.90 (10M) |
| GPT-6.1 Sol | $36.00 | $360.00 | $28.40 (10M) | $18.00 (10M) |
| GPT-6 Astra | $180.00 | $1,800.00 | $144.00 (10M) | $90.00 (10M) |
At these ratios, Astra costs 5x GPT-6.1 Sol, and GPT-6.1 Sol costs 20x Luna.
Long prompts cost more
Once a request’s input goes over 272K tokens, OpenAI bills it at long-context rates:
| Model | Input ≤272K / >272K | Output ≤272K / >272K |
|---|---|---|
| GPT-6 Luna | $0.1 / $0.2 | $0.5 / $0.75 |
| GPT-6.1 Sol | $2 / $4 | $10 / $15 |
| GPT-6 Astra | $10 / $20 | $50 / $75 |
Our calculator uses the short-context rates. Keep prompts under 272K tokens, or work out the long-context case by hand with the table above.
Speed tiers
OpenAI sells faster processing at a premium. In Fast mode, Astra is $20 / $100, GPT-6.1 Sol is $4 / $20, and Luna is $0.20 / $1.00. In Ultrafast, Astra is $60 / $300 and GPT-6.1 Sol is $12 / $60. Regional data-residency endpoints add 10%. All figures on this page are for the Standard tier.
Which tier to price first
- GPT-6 Luna costs the same per token as Claude Haiku 5.5, but it has no 100K-token cliff, and its long-context step starts at 272K.
- GPT-6.1 Sol costs the same per token as Claude Sonnet 5.5.
- GPT-6 Astra is the flagship tier and the most expensive per token.
Price your own traffic on the Astra, GPT-6.1 Sol or Luna pages. All OpenAI models are listed on the provider page.
Prices: OpenAI API pricing, verified 2026-10-25. Prices change; check the official page before committing budget.