Compare model prices
Input, output, cached and batch prices, every route and a month of tokens, side by side.
A month of tokens
GPT-6 Astra$2,000
GPT-6 Luna$20.00
100M input and 20M output tokens, list price. Use your own numbers.
Side by side
| GPT-6 Astra | GPT-6 Luna | |
|---|---|---|
| Input per 1M | $10 | $0.1 |
| Output per 1M | $50 | $0.5 |
| Cached input | $1 | $0.01 |
| Batch in / out | $5 / $25 | $0.05 / $0.25 |
| Blended (3:1) | $20 | $0.2 |
| Context | 1.05M | 1.05M |
| Routes | 6 | 7 |
| Released | Sep 4, 2026 | Sep 22, 2026 |
| Leadersboard score | 49 (#2) | n/a |
On the price map
Size: context window, 8K to 2MEqual blended cost (3:1) 243 models, lower left is cheaper
Every route
GPT-6 Astra
OpenAIflex-50%$5 / $25cached $0.5
Azure AI Foundrydefault region$10 / $50cached $1
OpenAI (list)the lab's own API$10 / $50cached $1
AWS Bedrockus-west-2+10%$11 / $55cached $1.1
Azure AI Foundryus+10%$11 / $55cached $1.1
OpenAIfast+100%$20 / $100cached $2
GPT-6 Luna
OpenAIflex-50%$0.05 / $0.25cached $0.005
Azure AI Foundrydefault region$0.1 / $0.5cached $0.01
OpenAI (list)the lab's own API$0.1 / $0.5cached $0.01
AWS Bedrockus-east-1+10%$0.11 / $0.55cached $0.011
Azure AI Foundryeu+10%$0.11 / $0.55cached $0.011
Azure AI Foundryus+10%$0.11 / $0.55cached $0.011
OpenAIfast+100%$0.2 / $1cached $0.02
List prices from providers' pages on the date shown; your contract may differ. Logos via logo.dev; trademarks belong to their owners.