GPT-6.2 Astra API Pricing: 4 Providers Compared
Input, cached input, cache write and output prices per million GPT-6.2 Astra tokens on OpenAI, OpenRouter, Vercel AI Gateway and ArgoLink for prompts of up to 272K tokens, read from each provider's own pages and APIs.

GPT-6.2 Astra costs $2.00 per million input tokens and $10.00 per million output tokens on ArgoLink, 75% below OpenAI's list price of $8.00 and $40.00. OpenRouter and Vercel AI Gateway charge OpenAI's list price, and OpenRouter adds a 5.5% fee when you buy credits. GPT-6.2 Astra costs 20% less per token than GPT-6 Astra on OpenAI and 13% less on ArgoLink.
We read the GPT-6.2 Astra price on OpenAI, OpenRouter and Vercel AI Gateway and set it next to ArgoLink's. The comparison covers each provider's input, cached input, cache write and output rates, what prompts over 272K tokens and prompt caching do to a GPT-6.2 Astra bill, and what real GPT-6.2 Astra jobs cost.
TL;DR
OpenAI describes GPT-6.2 Astra as "Our most capable model for your hardest projects, from start to finish." OpenAI charges $8.00 per million input tokens, $0.40 per million cached input tokens, $10.00 per million cache writes and $40.00 per million output tokens for GPT-6.2 Astra prompts of up to 272K tokens.
ArgoLink charges $2.00 input, $0.10 cached input, $2.50 cache writes and $10.00 output per million GPT-6.2 Astra tokens, 75% below OpenAI on each rate.
GPT-6.2 Astra costs 20% less than GPT-6 Astra on OpenAI, which lists GPT-6 Astra at $10.00 input and $50.00 output per million tokens. On ArgoLink, GPT-6.2 Astra costs 13% less than GPT-6 Astra, which costs $2.30 input and $11.50 output per million tokens.
OpenRouter and Vercel AI Gateway charge OpenAI's GPT-6.2 Astra list price, $8.00 input and $40.00 output per million tokens, and OpenRouter adds a 5.5% fee when you buy credits. Counting OpenRouter's credit fee, ArgoLink is 76% cheaper than OpenRouter and 75% cheaper than Vercel AI Gateway on GPT-6.2 Astra.
A GPT-6.2 Astra prompt over 272K tokens is billed at twice the input and cache rates and 1.5 times the output rate on OpenAI and on ArgoLink: $16.00 and $60.00 per million tokens on OpenAI, $4.00 and $15.00 on ArgoLink.
A month of 100 million input tokens and 10 million output tokens on GPT-6.2 Astra, in prompts of up to 272K tokens, costs $300.00 on ArgoLink, $1,200.00 on OpenAI and Vercel AI Gateway, and about $1,266.00 on OpenRouter with its credit fee.
How we checked: every competitor figure is the provider's own published GPT-6.2 Astra price, read from OpenAI's model page, OpenRouter's models API and pricing page, and Vercel's AI Gateway models API and pricing page.
How much does GPT-6.2 Astra cost per million tokens on each provider?
The table lists each provider's GPT-6.2 Astra price per million tokens for prompts of up to 272K tokens. OpenRouter's prices are before its 5.5% fee on credits.
| Per million tokens | Input | Cached input | Cache writes | Output | ArgoLink is cheaper by |
|---|---|---|---|---|---|
| ArgoLink | $2.00 | $0.10 | $2.50 | $10.00 | — |
| OpenAI (official) | $8.00 | $0.40 | $10.00 | $40.00 | 75% |
| OpenRouter | $8.00 | $0.40 | $10.00 | $40.00 | 75% |
| Vercel AI Gateway | $8.00 | $0.40 | $10.00 | $40.00 | 75% |
ArgoLink has the lowest GPT-6.2 Astra price of the four providers on all four token rates, for prompts of up to 272K tokens.
- GPT-6.2 Astra cached input costs one twentieth of the input rate on all four providers: $0.40 against $8.00 on OpenAI, and $0.10 against $2.00 on ArgoLink.
- GPT-6.2 Astra cache writes cost 1.25 times the input rate on all four providers: $10.00 on OpenAI and $2.50 on ArgoLink.
How much cheaper is GPT-6.2 Astra than GPT-6 Astra?
GPT-6.2 Astra lists at 20% less than GPT-6 Astra on OpenAI's input and output rates, and its cached input costs 60% less. The table lists GPT-6.2 Astra and GPT-6 Astra prices per million tokens for prompts of up to 272K tokens.
| Per million tokens | Input | Cached input | Cache writes | Output |
|---|---|---|---|---|
| GPT-6.2 Astra on ArgoLink | $2.00 | $0.10 | $2.50 | $10.00 |
| GPT-6 Astra on ArgoLink | $2.30 | $0.23 | $2.875 | $11.50 |
| GPT-6.2 Astra on OpenAI | $8.00 | $0.40 | $10.00 | $40.00 |
| GPT-6 Astra on OpenAI | $10.00 | $1.00 | $12.50 | $50.00 |
On ArgoLink, GPT-6.2 Astra costs 13% less than GPT-6 Astra on input and output. A month of 100 million input and 10 million output tokens costs $300.00 on GPT-6.2 Astra and $345.00 on GPT-6 Astra at ArgoLink's prices.
GPT-6.2 Astra saves more from prompt caching than GPT-6 Astra, because GPT-6.2 Astra prices cached input at one twentieth of its input rate and GPT-6 Astra at one tenth. A million cached input tokens cost $0.10 on GPT-6.2 Astra and $0.23 on GPT-6 Astra at ArgoLink.
How much cheaper is GPT-6.2 Astra on ArgoLink than on other providers?
ArgoLink charges 75% less than OpenAI on every GPT-6.2 Astra rate, for prompts of up to 272K tokens. The table compares ArgoLink's GPT-6.2 Astra input and output prices with each other provider's.
| Compared with | ArgoLink cheaper on input | ArgoLink cheaper on output |
|---|---|---|
| OpenAI | 75% | 75% |
| OpenRouter, with its 5.5% credit fee | 76% | 76% |
| Vercel AI Gateway | 75% | 75% |
Where does each GPT-6.2 Astra price come from?
OpenAI: the official GPT-6.2 Astra list price
OpenAI publishes the GPT-6.2 Astra price on the model's page, per million text tokens: $8.00 input, $0.40 cached input, $10.00 cache writes and $40.00 output. OpenAI's listed GPT-6.2 Astra rates apply to prompts of up to 272K tokens.
OpenAI's GPT-6.2 Astra page adds a rule for long prompts: "For requests with more than 272K input tokens, input and cache rates double and output costs 1.5 times the short-context rate for the full request."
OpenRouter: OpenAI's GPT-6.2 Astra price plus a credit fee
OpenRouter's models API lists GPT-6.2 Astra at $8.00 per million input tokens, $0.40 cached input, $10.00 cache writes and $40.00 output, the same as OpenAI. OpenRouter's pricing page states: "Inference is billed at the provider's list price on every plan."
OpenRouter adds a fee on top of the GPT-6.2 Astra token price. OpenRouter's pricing page states: "OpenRouter's fee is charged when you buy credits, 5.5% on Standard and 8% on Business, never on individual requests." On a Standard OpenRouter account, the fee makes a million GPT-6.2 Astra input tokens cost about $8.44 and a million output tokens about $42.20.
Vercel AI Gateway: OpenAI's GPT-6.2 Astra price with no markup
Vercel's AI Gateway models API lists GPT-6.2 Astra at $8.00 per million input tokens, $0.40 cached input, $10.00 cache writes and $40.00 output, the same as OpenAI. Vercel's pricing page states: "AI Gateway charges no markup and no platform fee on tokens. You pay the provider's list price on a pay-as-you-go basis."
ArgoLink: one discount on every GPT-6.2 Astra rate
ArgoLink charges 75% less than OpenAI's list price on all four GPT-6.2 Astra rates: $2.00 input, $0.10 cached input, $2.50 cache writes and $10.00 output per million tokens.
ArgoLink follows OpenAI's GPT-6.2 Astra billing rules on its own price. On ArgoLink, GPT-6.2 Astra prompts over 272K tokens cost twice the input and cache rates and 1.5 times the output rate, and GPT-6.2 Astra reasoning tokens are billed as output tokens.
ArgoLink uses OpenAI's model ID, gpt-6.2-astra, and OpenAI's Responses API format. The GPT-6.2 Astra model page shows the price a request is billed at.
What does a GPT-6.2 Astra job cost on each provider?
The table prices three GPT-6.2 Astra jobs at each provider's rates for prompts of up to 272K tokens, without prompt caching. Output includes reasoning tokens, and the OpenRouter column includes its 5.5% credit fee.
| GPT-6.2 Astra job | ArgoLink | OpenAI (official) | Vercel AI Gateway | OpenRouter, with its fee |
|---|---|---|---|---|
| 1,000 chat requests of 2,000 input and 1,000 output tokens each | $14.00 | $56.00 | $56.00 | $59.08 |
| 100 document reviews of 200,000 input and 4,000 output tokens each | $44.00 | $176.00 | $176.00 | $185.68 |
| A month of 100 million input and 10 million output tokens | $300.00 | $1,200.00 | $1,200.00 | $1,266.00 |
OpenRouter's amounts are approximate: they add its 5.5% credit fee to its token prices.
A month of 100 million GPT-6.2 Astra input tokens and 10 million output tokens costs $900.00 less on ArgoLink than on OpenAI.
How much do GPT-6.2 Astra prompts over 272K tokens cost?
GPT-6.2 Astra costs more per token once a prompt passes 272K input tokens. OpenAI's GPT-6.2 Astra page states: "For requests with more than 272K input tokens, input and cache rates double and output costs 1.5 times the short-context rate for the full request." ArgoLink applies the same multipliers to its own GPT-6.2 Astra price.
The table gives GPT-6.2 Astra's price per million tokens on ArgoLink and OpenAI below and above the 272K threshold.
| Per million tokens | ArgoLink, prompts up to 272K | ArgoLink, prompts over 272K | OpenAI, prompts up to 272K | OpenAI, prompts over 272K |
|---|---|---|---|---|
| Input | $2.00 | $4.00 | $8.00 | $16.00 |
| Cached input | $0.10 | $0.20 | $0.40 | $0.80 |
| Cache writes | $2.50 | $5.00 | $10.00 | $20.00 |
| Output | $10.00 | $15.00 | $40.00 | $60.00 |
The higher GPT-6.2 Astra rate applies to the whole request, not only to the tokens past 272K. On ArgoLink, cached tokens count toward the 272K threshold.
- A GPT-6.2 Astra request with a 270,000-token prompt and 10,000 output tokens costs $0.64 on ArgoLink and $2.56 on OpenAI.
- A GPT-6.2 Astra request with a 280,000-token prompt and 10,000 output tokens costs $1.27 on ArgoLink and $5.08 on OpenAI, about 2 times the 270,000-token request.
GPT-6.2 Astra accepts a context of up to 1,050,000 tokens, of which up to 922,000 can be input, and returns up to 128,000 output tokens, according to OpenAI's GPT-6.2 Astra page. Splitting a long job into prompts of up to 272K tokens keeps each GPT-6.2 Astra request at the lower rate.
How much does prompt caching save on GPT-6.2 Astra?
GPT-6.2 Astra cached input costs one twentieth of the input rate, and a cache write costs 1.25 times the input rate, on all four providers. Prompt caching pays off for a GPT-6.2 Astra prompt that is sent again and again, such as a long system prompt, a shared document or an agent's history.
The table prices one 100,000-token GPT-6.2 Astra prompt sent 100 times at each provider's rates: once without caching, and once written to the cache on the first request and read from the cache on the other 99.
| Provider | 100 sends without caching | Written once, read 99 times from the cache |
|---|---|---|
| ArgoLink | $20.00 | $1.24 |
| OpenAI (official) | $80.00 | $4.96 |
| Vercel AI Gateway | $80.00 | $4.96 |
| OpenRouter, with its fee | $84.40 | $5.23 |
With caching, a 100,000-token GPT-6.2 Astra prompt sent 100 times costs $1.24 on ArgoLink and $4.96 on OpenAI, 94% less than without caching on every provider in the table.
A GPT-6.2 Astra cache write keeps the prompt for 30 minutes. OpenAI's GPT-6.2 Astra page states: "The cache write price is for 30 minutes of retention." A GPT-6.2 Astra prompt sent again within 30 minutes is read from the cache at the cached input rate.
How are GPT-6.2 Astra reasoning tokens billed?
GPT-6.2 Astra reasons before it answers, and its reasoning tokens are billed as output tokens. OpenAI's reasoning guide states: "While reasoning tokens are not visible via the API, they still occupy space in the model's context window and are billed as output tokens."
ArgoLink bills GPT-6.2 Astra reasoning tokens as output tokens too, at $10.00 per million output tokens, against $40.00 on OpenAI.
GPT-6.2 Astra does not accept a reasoning effort of none. OpenAI's GPT-6.2 Astra page states: "The API rejects requests that set reasoning effort to none." OpenAI's GPT-6.2 migration guide advises: "Start with medium for a new integration. If you currently use none or minimal, use low instead." ArgoLink accepts GPT-6.2 Astra requests at the low, medium, high, xhigh and max reasoning efforts.
OpenAI's reasoning guide names max_output_tokens as the cap on every token a reasoning model generates, reasoning tokens included. A GPT-6.2 Astra request with max_output_tokens set to 8,000 is billed for at most 8,000 output tokens: $0.08 on ArgoLink and $0.32 on OpenAI.
How does GPT-6.2 Astra's price compare with GPT-6.2 Sol?
GPT-6.2 Astra costs 4 times as much as GPT-6.2 Sol per input and output token, on ArgoLink and on OpenAI. The table lists both GPT-6.2 models' ArgoLink price next to OpenAI's list price per million tokens.
| Per million tokens | Provider | Input | Cached input | Output |
|---|---|---|---|---|
| GPT-6.2 Astra | ArgoLink | $2.00 | $0.10 | $10.00 |
| GPT-6.2 Astra | OpenAI (official) | $8.00 | $0.40 | $40.00 |
| GPT-6.2 Sol | ArgoLink | $0.50 | $0.025 | $2.50 |
| GPT-6.2 Sol | OpenAI (official) | $2.00 | $0.10 | $10.00 |
OpenAI's changelog advises: "Use GPT-6.2 Astra for demanding reasoning and coding tasks, or GPT-6.2 Sol for everyday coding and agent workflows." GPT-6.2 Sol costs $0.50 per million input tokens and $2.50 per million output tokens on ArgoLink.
The GPT-6.2 Sol pricing guide compares GPT-6.2 Sol on the same four providers, the GPT-6 Astra pricing guide does the same for GPT-6 Astra, the GPT API pricing comparison covers more GPT models, and the GPT-6.2 Astra guide covers the model itself.
Which GPT-6.2 Astra provider should you choose?
Choose by where the rest of your stack runs and by what GPT-6.2 Astra tokens cost you.
Choose OpenAI if you want to buy GPT-6.2 Astra directly from the company that builds it, at the list price.
Choose OpenRouter if you already route other models through OpenRouter and want one bill. You pay OpenAI's GPT-6.2 Astra price plus the 5.5% credit fee.
Choose Vercel AI Gateway if your application already runs on Vercel. Vercel's pricing page states: "AI Gateway charges no markup and no platform fee on tokens. You pay the provider's list price on a pay-as-you-go basis."
Choose ArgoLink if price per token decides: ArgoLink's GPT-6.2 Astra price is 75% below OpenAI, OpenRouter and Vercel AI Gateway on every token rate, at prompts of every length.
How were these GPT-6.2 Astra prices checked?
Each competitor price is the provider's own published GPT-6.2 Astra figure, read from the pages listed below: OpenAI's GPT-6.2 Astra page, changelog, migration guide and reasoning guide, OpenRouter's models API and pricing page, and Vercel's AI Gateway models API and pricing page. Every price is per million tokens for prompts of up to 272K tokens unless the article says otherwise.
OpenRouter's fee-adjusted GPT-6.2 Astra figures add the 5.5% fee that OpenRouter charges on credit purchases on a Standard account.
ArgoLink's prices come from the catalog that the ArgoLink model page and ArgoLink's billing both read, so the GPT-6.2 Astra price in this article is the price a request is charged.
We are ArgoLink, one of the four providers compared. Prices change. If a figure here is out of date, the provider's own page is right.
- OpenAI: API changelog, GPT-6.2 Astra and GPT-6.2 Sol release entry. Checked November 18, 2026.
- OpenAI: GPT-6.2 Astra model page, text token prices, limits and reasoning settings. Checked November 18, 2026.
- OpenAI: Reasoning models guide, how reasoning tokens are billed. Checked November 18, 2026.
- OpenAI: Using GPT-6.2, migration checklist. Checked November 18, 2026.
- OpenRouter: Models API, price of openai/gpt-6.2-astra. The API states USD per token; the article states USD per million tokens. Checked November 18, 2026.
- OpenRouter: Pricing page, plans and platform fees. Checked November 18, 2026.
- Vercel: AI Gateway models API, price of openai/gpt-6.2-astra. The API states USD per token; the article states USD per million tokens. Checked November 18, 2026.
- Vercel: AI Gateway pricing. Checked November 18, 2026.
Get started with GPT-6.2 Astra on ArgoLink
GPT-6.2 Astra on ArgoLink works with the OpenAI API at one base URL, https://argolink.io/v1, and an ArgoLink API key. The model ID is gpt-6.2-astra, the same as OpenAI's.
curl -sS https://argolink.io/v1/chat/completions \
-H "Authorization: Bearer $ARGOLINK_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-6.2-astra",
"messages": [
{"role": "user", "content": "Explain what this model is good at in one sentence."}
]
}'
The OpenAI Python SDK calls GPT-6.2 Astra on ArgoLink with the base URL set to https://argolink.io/v1:
import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ["ARGOLINK_API_KEY"],
base_url="https://argolink.io/v1",
)
response = client.chat.completions.create(
model="gpt-6.2-astra",
messages=[
{"role": "user", "content": "Explain what this model is good at in one sentence."}
],
)
print(response.choices[0].message.content)
GPT-6.2 Astra calls tools only through the Responses API. OpenAI's GPT-6.2 migration guide states: "GPT-6.2 Astra and GPT-6.2 Sol support Chat Completions, but tool calling requires Responses." ArgoLink serves GPT-6.2 Astra on both /v1/responses and /v1/chat/completions under the same rule.
The GPT-6.2 Astra model page lists every parameter and the live price. The API guide covers keys and endpoints, and Models & Pricing lists every model ArgoLink serves.
GPT-6.2 Astra API pricing FAQs
What is the cheapest GPT-6.2 Astra API?
ArgoLink is the cheapest of the four providers compared here. GPT-6.2 Astra costs $2.00 per million input tokens and $10.00 per million output tokens on ArgoLink, 75% below OpenAI, OpenRouter and Vercel AI Gateway.
How much does GPT-6.2 Astra cost per million tokens?
On OpenAI, GPT-6.2 Astra costs $8.00 per million input tokens, $0.40 per million cached input tokens, $10.00 per million cache writes and $40.00 per million output tokens. On ArgoLink, GPT-6.2 Astra costs $2.00, $0.10, $2.50 and $10.00.
Do GPT-6.2 Astra prompts over 272K tokens cost more?
Yes. OpenAI bills a GPT-6.2 Astra prompt over 272K input tokens at twice the input and cache rates and 1.5 times the output rate, for the whole request: $16.00 input and $60.00 output per million tokens. ArgoLink applies OpenAI's 272K rule to its own price, so a GPT-6.2 Astra prompt over 272K tokens costs $4.00 input and $15.00 output per million tokens on ArgoLink.
Can GPT-6.2 Astra call tools through Chat Completions?
No. OpenAI's GPT-6.2 migration guide states: "GPT-6.2 Astra and GPT-6.2 Sol support Chat Completions, but tool calling requires Responses." ArgoLink follows the same rule: send GPT-6.2 Astra requests with tools to /v1/responses.
Is GPT-6.2 Astra cheaper than GPT-6 Astra?
Yes. GPT-6.2 Astra lists at $8.00 input and $40.00 output per million tokens on OpenAI, against $10.00 and $50.00 for GPT-6 Astra. On ArgoLink, GPT-6.2 Astra costs $2.00 and $10.00, against $2.30 and $11.50 for GPT-6 Astra.
What is GPT-6.2 Astra's context window?
GPT-6.2 Astra has a context window of 1,050,000 tokens, takes up to 922,000 input tokens and returns up to 128,000 output tokens per request, according to OpenAI's GPT-6.2 Astra page. GPT-6.2 Astra supports text and image input with text output, and its knowledge cutoff is May 18, 2026, according to OpenAI's GPT-6.2 Astra page.