Verified 2026-10-07 · sourced from OpenAI
GPT-6 Astra Token Calculator, Pricing & 100K/1M Cost
Check OpenAI GPT-6 Astra pricing, estimate 100K and 1M token cost, and size a real API budget before you send a single request. Standard pricing is $10.00 per million input tokens and $50.00 per million output tokens with a 1.1M token context window.
Quick answer: GPT-6 Astra pricing per 1M tokens is $10.00 input and $50.00 output. Context window: 1,050,000 tokens · Cached input: $1.000 / 1M.
Best for searches like GPT-6 Astra token calculator, GPT-6 Astra pricing, GPT-6 Astra 100K tokens cost, GPT-6 Astra 1M token cost.
Reference rates: Standard · ≤272K input · global
USD per million text tokens. Verified 2026-10-07 · Official source
| Mode / prompt size | Input | Cache read | Cache write | Output |
|---|---|---|---|---|
| Standard ≤272,000 input | $10 | $1 | $12.5 | $50 |
| Standard >272,000 input | $20 | $2 | $25 | $75 |
| Batch ≤272,000 input | $5 | $0.5 | $6.25 | $25 |
| Batch >272,000 input | $10 | $1 | $12.5 | $37.5 |
| Flex ≤272,000 input | $5 | $0.5 | $6.25 | $25 |
| Flex >272,000 input | $10 | $1 | $12.5 | $37.5 |
| Fast mode ≤272,000 input | $20 | $2 | $25 | $100 |
| Fast mode >272,000 input | $40 | $4 | $50 | $150 |
| Ultrafast ≤272,000 input | $60 | $6 | $75 | $300 |
| Ultrafast >272,000 input | $120 | $12 | $150 | $450 |
1,050,000 context tokens · max output 128,000. Local token counts are estimates; use the provider’s usage counts for an invoice estimate.
Text-only estimate. Local token counts are approximate; use provider-reported usage for billing. Tool, image, audio and video charges are excluded.
Above 272,000 input tokens, the full request uses 2× input/cache rates and 1.5× output rates. Batch/Flex and Fast/Ultrafast are separate processing options, not stackable discounts.
Global endpoint rates shown. Regional processing adds 10% where available and is excluded from this estimate.
Billed output includes reasoning/thinking tokens that cannot be inferred from pasted final text; use API-reported output usage for accurate billing.
Pick the route that matches what you searched for
Some visitors want a fast GPT-6 Astra API cost estimate, others want a direct 100K or 1M token budget, and some are already comparing alternatives. These shortcuts remove the extra click.
Estimate a single request or prompt budget right now.
Jump straight to the most common budgeting checkpoint.
Use this when you are sizing production traffic or a monthly plan.
Open the closest head-to-head comparison instead of researching from scratch.
Context window
1,050,000 tokens
Input price
$10.00 / 1M
Output price
$50.00 / 1M
Cached input
$1.000 / 1M
Pricing modes and thresholds
Long-context pricing starts above 272,000 input tokens.
Usage scenarios
Compare standard and cached pricing (where available) across common workloads.
| Scenario | Tokens in | Tokens out | Total tokens | Standard cost | Cached cost |
|---|---|---|---|---|---|
Quick chat reply Single user question with a short assistant answer | 650 | 220 | 870 | $0.0175 | $0.0117 |
Coding assistant session Multi-turn pair programming exchange (≈6 turns) | 2,600 | 1,400 | 4,000 | $0.0960 | $0.0726 |
Knowledge base response Retrieval-augmented answer referencing multiple passages | 12,000 | 3,000 | 15,000 | $0.270 | $0.162 |
Near-max context run Large document processing approaching the 1.1M token limit | 924,000 | 126,000 | 1,050,000 | $27.93 | $11.30 |
Daily & monthly budgeting
Translate usage into predictable operating expenses across popular deployment sizes.
| Profile | Requests/day | Tokens/day | Daily cost | Monthly cost | Cached daily | Cached monthly |
|---|---|---|---|---|---|---|
| Team pilot | 25 | 75,000 | $1.75 | $52.50 | $1.30 | $39.00 |
| Product launch | 100 | 500,000 | $18.25 | $547.50 | $11.95 | $358.50 |
| Enterprise scale | 500 | 3,000,000 | $115.00 | $3450.00 | $79.00 | $2370.00 |
Pricing notes
- Text-only estimate. Local token counts are approximate; use provider-reported usage for billing. Tool, image, audio and video charges are excluded.
- Above 272,000 input tokens, the full request uses 2× input/cache rates and 1.5× output rates. Batch/Flex and Fast/Ultrafast are separate processing options, not stackable discounts.
- Global endpoint rates shown. Regional processing adds 10% where available and is excluded from this estimate.
- Billed output includes reasoning/thinking tokens that cannot be inferred from pasted final text; use API-reported output usage for accurate billing.
Frequently asked questions
How much does GPT-6 Astra cost per 1,000 tokens?
At the published rates of $10.00 per million input tokens and $50.00 per million output tokens, a typical 1,000 token request (≈70% input, 30% output) costs about $0.0220.
Does GPT-6 Astra offer cached input discounts?
GPT-6 Astra drops input costs to $1.000 per million cached tokens. Using cached contexts, that same 1,000 token call totals $0.0157, a significant saving for chatbots and RAG systems.
What is the context window for GPT-6 Astra?
GPT-6 Astra supports up to 1,050,000 tokens (1.1M), allowing large prompts and retrieval-augmented payloads in a single call.
How fresh is the GPT-6 Astra pricing data?
Pricing is sourced from https://developers.openai.com/api/docs/models/gpt-6-astra and was last verified on 2026-10-07. The calculator updates automatically when models.json is refreshed.