Seedance 2.5 — 15% below official pricing, 57% cheaper than Higgsfield and Fal. Draft mode is live · Wan 3.0 — 35% below official list price.

Guides, updates and engineering notes from ArgoLink on building with one API for every AI model. Blog

GPT-6.2 Sol API Pricing: 4 Providers Compared

Input, cached input, cache write and output prices per million GPT-6.2 Sol 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.

Bar chart: GPT-6.2 Sol API price per million input tokens. Prompts of up to 272K tokens. Each provider's own published price. ArgoLink $0.50, OpenAI (official) $2.00, Vercel AI Gateway $2.00, OpenRouter $2.00.

GPT-6.2 Sol costs $0.50 per million input tokens and $2.50 per million output tokens on ArgoLink, 75% below OpenAI's list price of $2.00 and $10.00. OpenRouter and Vercel AI Gateway charge OpenAI's list price, and OpenRouter adds a 5.5% fee when you buy credits. A GPT-6.2 Sol prompt over 272K tokens costs twice the input and cache rates and 1.5 times the output rate on OpenAI and on ArgoLink.

We read the GPT-6.2 Sol 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 Sol bill, and what real GPT-6.2 Sol jobs cost.

TL;DR

OpenAI describes GPT-6.2 Sol as "Our go-to model for everyday work, from coding to powering agents at scale." OpenAI charges $2.00 per million input tokens, $0.10 per million cached input tokens, $2.50 per million cache writes and $10.00 per million output tokens for GPT-6.2 Sol prompts of up to 272K tokens.

ArgoLink charges $0.50 input, $0.025 cached input, $0.625 cache writes and $2.50 output per million GPT-6.2 Sol tokens, 75% below OpenAI on each rate.

OpenRouter and Vercel AI Gateway charge OpenAI's GPT-6.2 Sol list price, $2.00 input and $10.00 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 Sol.

A GPT-6.2 Sol 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: $4.00 and $15.00 per million tokens on OpenAI, $1.00 and $3.75 on ArgoLink.

A month of 100 million input tokens and 10 million output tokens on GPT-6.2 Sol, in prompts of up to 272K tokens, costs $75.00 on ArgoLink, $300.00 on OpenAI and Vercel AI Gateway, and about $316.50 on OpenRouter with its credit fee.

How we checked: every competitor figure is the provider's own published GPT-6.2 Sol 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 Sol cost per million tokens on each provider?

The table lists each provider's GPT-6.2 Sol 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 $0.50 $0.025 $0.625 $2.50 —
OpenAI (official) $2.00 $0.10 $2.50 $10.00 75%
OpenRouter $2.00 $0.10 $2.50 $10.00 75%
Vercel AI Gateway $2.00 $0.10 $2.50 $10.00 75%

ArgoLink has the lowest GPT-6.2 Sol price of the four providers on all four token rates, for prompts of up to 272K tokens. OpenRouter and Vercel AI Gateway match OpenAI's list price, and OpenRouter's credit fee puts what you pay for GPT-6.2 Sol through OpenRouter above OpenAI's price.

  • GPT-6.2 Sol cached input costs one twentieth of the input rate on all four providers: $0.10 against $2.00 on OpenAI, and $0.025 against $0.50 on ArgoLink.
  • GPT-6.2 Sol cache writes cost 1.25 times the input rate on all four providers: $2.50 on OpenAI and $0.625 on ArgoLink.

ArgoLink charges 75% less than OpenAI on every GPT-6.2 Sol rate, for prompts of up to 272K tokens. The table compares ArgoLink's GPT-6.2 Sol 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%

ArgoLink's GPT-6.2 Sol cached input and cache write prices are 75% below OpenAI, like its input price, because every provider here prices cached input at one twentieth and cache writes at 1.25 times its own input rate. A million cached GPT-6.2 Sol input tokens cost $0.025 on ArgoLink and $0.10 on OpenAI.

Where does each GPT-6.2 Sol price come from?

OpenAI: the official GPT-6.2 Sol list price

OpenAI publishes the GPT-6.2 Sol price on the model's page, per million text tokens: $2.00 input, $0.10 cached input, $2.50 cache writes and $10.00 output. OpenAI's listed GPT-6.2 Sol rates apply to prompts of up to 272K tokens.

OpenAI's GPT-6.2 Sol 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 Sol price plus a credit fee

OpenRouter's models API lists GPT-6.2 Sol at $2.00 per million input tokens, $0.10 cached input, $2.50 cache writes and $10.00 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 Sol 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 Sol input tokens cost about $2.11 and a million output tokens about $10.55.

Vercel AI Gateway: OpenAI's GPT-6.2 Sol price with no markup

Vercel's AI Gateway models API lists GPT-6.2 Sol at $2.00 per million input tokens, $0.10 cached input, $2.50 cache writes and $10.00 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 charges 75% less than OpenAI's list price on all four GPT-6.2 Sol rates: $0.50 input, $0.025 cached input, $0.625 cache writes and $2.50 output per million tokens.

ArgoLink follows OpenAI's GPT-6.2 Sol billing rules on its own price. On ArgoLink, GPT-6.2 Sol prompts over 272K tokens cost twice the input and cache rates and 1.5 times the output rate, and GPT-6.2 Sol reasoning tokens are billed as output tokens.

ArgoLink uses OpenAI's model ID, gpt-6.2-sol, and OpenAI's Responses API format. The GPT-6.2 Sol model page shows the price a request is billed at.

What does a GPT-6.2 Sol job cost on each provider?

The table prices three GPT-6.2 Sol 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 Sol 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 $3.50 $14.00 $14.00 $14.77
100 document reviews of 200,000 input and 4,000 output tokens each $11.00 $44.00 $44.00 $46.42
A month of 100 million input and 10 million output tokens $75.00 $300.00 $300.00 $316.50

OpenRouter's amounts are approximate: they add its 5.5% credit fee to its token prices.

A month of 100 million GPT-6.2 Sol input tokens and 10 million output tokens costs $225.00 less on ArgoLink than on OpenAI.

The Codex CLI setup guide shows how to point Codex at ArgoLink, so a coding agent's GPT-6.2 Sol tokens are billed at ArgoLink's price.

How much do GPT-6.2 Sol prompts over 272K tokens cost?

GPT-6.2 Sol costs more per token once a prompt passes 272K input tokens. OpenAI's GPT-6.2 Sol 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 Sol price.

The table gives GPT-6.2 Sol'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 $0.50 $1.00 $2.00 $4.00
Cached input $0.025 $0.05 $0.10 $0.20
Cache writes $0.625 $1.25 $2.50 $5.00
Output $2.50 $3.75 $10.00 $15.00

The higher GPT-6.2 Sol 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 Sol request with a 270,000-token prompt and 10,000 output tokens costs $0.16 on ArgoLink and $0.64 on OpenAI.
  • A GPT-6.2 Sol request with a 280,000-token prompt and 10,000 output tokens costs $0.32 on ArgoLink and $1.27 on OpenAI, about 2 times the 270,000-token request.

GPT-6.2 Sol 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 Sol page. Splitting a long job into prompts of up to 272K tokens keeps each GPT-6.2 Sol request at the lower rate.

How much does prompt caching save on GPT-6.2 Sol?

GPT-6.2 Sol 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 Sol 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 Sol 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 $5.00 $0.31
OpenAI (official) $20.00 $1.24
Vercel AI Gateway $20.00 $1.24
OpenRouter, with its fee $21.10 $1.31

With caching, a 100,000-token GPT-6.2 Sol prompt sent 100 times costs $0.31 on ArgoLink and $1.24 on OpenAI, 94% less than without caching on every provider in the table.

A GPT-6.2 Sol cache write keeps the prompt for 30 minutes. OpenAI's GPT-6.2 Sol page states: "The cache write price is for 30 minutes of retention." A GPT-6.2 Sol prompt sent again within 30 minutes is read from the cache at the cached input rate.

How are GPT-6.2 Sol reasoning tokens billed?

GPT-6.2 Sol 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 Sol reasoning tokens as output tokens too, at $2.50 per million output tokens, against $10.00 on OpenAI.

GPT-6.2 Sol does not accept a reasoning effort of none. OpenAI's GPT-6.2 Sol 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 Sol 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 Sol request with max_output_tokens set to 8,000 is billed for at most 8,000 output tokens: $0.02 on ArgoLink and $0.08 on OpenAI.

How does GPT-6.2 Sol's price compare with GPT-6.2 Astra and GPT-6 Astra?

GPT-6.2 Sol costs less than GPT-6.2 Astra and GPT-6 Astra on both OpenAI and ArgoLink. The table lists each model's ArgoLink price next to OpenAI's list price per million tokens.

Per million tokens Provider Input Cached input Output
GPT-6.2 Sol ArgoLink $0.50 $0.025 $2.50
GPT-6.2 Sol OpenAI (official) $2.00 $0.10 $10.00
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 Astra ArgoLink $2.30 $0.23 $11.50
GPT-6 Astra OpenAI (official) $10.00 $1.00 $50.00

GPT-6.2 Astra costs 4 times as much as GPT-6.2 Sol per input and output token, on ArgoLink and on OpenAI. 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 Astra costs $2.30 per million input tokens and $11.50 per million output tokens on ArgoLink, against $0.50 and $2.50 for GPT-6.2 Sol.

The GPT-6.2 Astra pricing guide compares GPT-6.2 Astra on the same four providers, the GPT API pricing comparison covers more GPT models, and the GPT-6.2 Sol guide covers the model itself.

Which GPT-6.2 Sol provider should you choose?

Choose by where the rest of your stack runs and by what GPT-6.2 Sol tokens cost you.

Choose OpenAI if you want to buy GPT-6.2 Sol 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 Sol 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 Sol 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 Sol prices checked?

Each competitor price is the provider's own published GPT-6.2 Sol figure, read from the pages listed below: OpenAI's GPT-6.2 Sol 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 Sol 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 Sol 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 Sol 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-sol. 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-sol. 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.

GPT-6.2 Sol 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-sol, 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-sol",
    "messages": [
      {"role": "user", "content": "Explain what this model is good at in one sentence."}
    ]
  }'

The OpenAI Python SDK calls GPT-6.2 Sol 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-sol",
    messages=[
        {"role": "user", "content": "Explain what this model is good at in one sentence."}
    ],
)
print(response.choices[0].message.content)

A GPT-6.2 Sol request with 1,000 input tokens and 500 output tokens costs $0.0018 on ArgoLink.

GPT-6.2 Sol 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 Sol on both /v1/responses and /v1/chat/completions under the same rule.

The GPT-6.2 Sol 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 Sol API pricing FAQs

What is the cheapest GPT-6.2 Sol API?

ArgoLink is the cheapest of the four providers compared here. GPT-6.2 Sol costs $0.50 per million input tokens and $2.50 per million output tokens on ArgoLink, 75% below OpenAI, OpenRouter and Vercel AI Gateway.

How much does GPT-6.2 Sol cost per million tokens?

On OpenAI, GPT-6.2 Sol costs $2.00 per million input tokens, $0.10 per million cached input tokens, $2.50 per million cache writes and $10.00 per million output tokens. On ArgoLink, GPT-6.2 Sol costs $0.50, $0.025, $0.625 and $2.50.

Is GPT-6.2 Sol cheaper on OpenRouter or Vercel AI Gateway than on OpenAI?

No. OpenRouter and Vercel AI Gateway both charge OpenAI's GPT-6.2 Sol list price, $2.00 input and $10.00 output per million tokens. OpenRouter adds a 5.5% fee when you buy credits on a Standard account, so GPT-6.2 Sol usage costs about 5.5% more through OpenRouter than on OpenAI.

Do GPT-6.2 Sol prompts over 272K tokens cost more?

Yes. OpenAI bills a GPT-6.2 Sol prompt over 272K input tokens at twice the input and cache rates and 1.5 times the output rate, for the whole request: $4.00 input and $15.00 output per million tokens. ArgoLink applies OpenAI's 272K rule to its own price, so a GPT-6.2 Sol prompt over 272K tokens costs $1.00 input and $3.75 output per million tokens on ArgoLink.

Can GPT-6.2 Sol 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 Sol requests with tools to /v1/responses.

Which reasoning efforts does GPT-6.2 Sol accept?

GPT-6.2 Sol accepts the low, medium, high, xhigh and max reasoning efforts on ArgoLink and rejects none, as OpenAI's API does. 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."

What is GPT-6.2 Sol's context window?

GPT-6.2 Sol 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 Sol page. OpenAI's GPT-6.2 Sol page states: "GPT-6.2 Sol accepts text and image inputs and produces text." The knowledge cutoff of GPT-6.2 Sol is June 1, 2026.

Welcome Builders

Who runs ArgoLink, refunds and how your data is handled

Request Refund

Submit a request and review its status in Top-up Orders.