AiCostCompare

GPT Astra Latest

by ~openai · ~openai/gpt-astra-latest · #406 cheapest of 422 paid models

Prices updated Sep 19, 2026, 6:22 PM UTC · refreshed hourly

This model always redirects to the latest model in the GPT Astra family.

Compare vs:Claude Sonnet 5Claude Opus 4.6Claude Haiku 4.5GPT-5.6 Sol

Pricing

Input

$10.00

per 1M tokens

Output

$50.00

per 1M tokens

Blended (3:1)

$20.00

3 input : 1 output

Cache read

$1.00

per 1M cached tokens

Cache write

$12.50

per 1M tokens

Cache & batch economics

Effective prices for GPT Astra Latest with prompt caching.

Cache hit rateEffective blended $/1MExample request*vs no cache
0%$20.00$1.07
50%$16.63$0.62−42%
90%$13.93$0.26−76%

*Example: 100K sticky context + 2K new input + 1K output. Batch discount is an approximate 50% off list rates (available for OpenAI / Anthropic / Google — not flagged for this creator). Confirm on the vendor’s pricing page.

What a request costs

WorkloadInput tokensOutput tokensCost / requestCost / 1K requests
Short chat message500300$0.02$20.00
Document summary8,0001,000$0.13$130.00
Codebase question (RAG)30,0002,000$0.4$400.00
Long-context analysis150,0005,000$1.75$1,750

Specs

Context window

1.05M

tokens

Max output

128K

tokens

Input modalities

file, image, text

Tokenizer

Router