OpenAI API Pricing in 2026 — How Much GPT Costs per Request, Model by Model

The OpenAI API is billed per token, not per month. Input and output have different prices, and output is the expensive part. On Kunavo, the range runs from $0.04 / $0.20 per million tokens with GPT-6 Luna to $4.00 / $20.00 with GPT-6 Astra, OpenAI’s top-end model; in between are GPT-6 Sol, at $0.80 / $4.00, and the workhorse GPT-5.6 Terra, at $0.70 / $4.20 — between 60% and 65% below OpenAI's rate, depending on the model. A ChatGPT subscription is a separate bill: it does not include the API, and the API does not require a subscription.

Prices checked on October 3, 2026. This page reads Kunavo prices directly from the model catalog: what you see is what the API charges. The “OpenAI pricing” column is the price OpenAI publishes for the same model at openai.com/api/pricing. Kunavo is a gateway that resells access to the models, not an independent comparison service.

OpenAI API prices by model, available to call today

Many Spanish pricing tables still list GPT-4o and its variants. This one includes only what a key can call today, in dollars per million tokens, as charged by Kunavo:

Live prices are always on the pricing page and on each model page, such as GPT-6 Sol or GPT-6 Luna. Which one to use for what:

• gpt-6-luna — according to OpenAI, its most efficient model for focused, high-volume tasks such as classification and extraction. Kunavo’s cheapest GPT.
• gpt-5-6-terra — the workhorse of the 5.6 family: RAG, assistants, and most production tasks.
• gpt-6-sol — OpenAI positions it for complex code and agent workflows. On Kunavo, $0.80 / $4.00: below GPT-5.6 Terra on output, slightly higher on input. The sensible default for code and agents.
• gpt-6-astra — the top end: cutting-edge reasoning and long agent sessions. For the difficult minority of tasks.
• gpt-5-6-sol and gpt-5-5 — previous generations. At Kunavo, they cost the same per token; moving from gpt-5-5 to the next generation requires changing only the model name.

GPT-6 Sol and GPT-6 Luna share a context window of 1,050,000 tokens and return up to 128,000 output tokens. A surcharge for very long requests: above 272,000 prompt tokens, GPT-5.5, the GPT-5.6 family, and the three GPT-6 models — Astra, Sol, and Luna — charge the entire request at 2× the input rate and 1.5× the output rate.

ChatGPT subscription and OpenAI API: two products, two bills

People searching for “ChatGPT API price” often want to know something else first: whether the subscription they already pay for covers the API. It does not: a subscription does not include an API key or balance, and API balance does not expand ChatGPT limits. Plan prices are listed at openai.com/chatgpt/pricing and change often, so they are not repeated here. For someone who spends all day in the chat window, the subscription is often cheaper: a turn with 2,000 input tokens and 700 output tokens costs about $0.0044 with GPT-6 Sol and $0.00434 with GPT-5.6 Terra; divide the plan price by that figure and you have the break-even point. The API wins for software, automation, and irregular use, where a quiet month should cost nothing.

What a request really costs

At Kunavo prices, without cache and without reasoning tokens — in other words, the floor of what appears on the bill:

With the minimum $10 top-up, you get about 1590 RAG responses on GPT-5.6 Terra, or about 29,400 on GPT-6 Luna. Two things every estimate overlooks: reasoning tokens, which are billed at the output rate even when they do not appear in the text, and retries: a response that arrived successfully and your code discarded has still been paid for. Requests that fail on Kunavo are not billed. To save money, prompt caching bills reused tokens at a fraction of the input rate: $0.20 instead of $2.00 per million with GPT-5.6 Sol (explained in the caching documentation).

Paying for the API from Spain or Latin America

Kunavo charges through Stripe, with prepaid balance in dollars: no subscription and no need to configure an account on the OpenAI platform. At checkout:

• Visa, Mastercard, American Express, JCB, and UnionPay cards, credit or debit
• Link, Apple Pay, and Google Pay

Stripe shows the amount in your currency when you pay. The minimum top-up is $10; larger top-ups receive a bonus: $100 converts to $110.00, $1000 converts to $1200.00, and $5000 converts to $6250.00. The balance does not expire. If what you are specifically looking for is avoiding a card, the steps are in the API without a credit card.

Move your OpenAI code to Kunavo

Kunavo is OpenAI-compatible: the base_url and the key change, nothing else. The response includes the usual usage field, which can be used to verify each charge.

from openai import OpenAI

client = OpenAI(
    api_key="sk-kn-...",
    base_url="https://api.kunavo.com/v1",  # en lugar de api.openai.com
)

resp = client.chat.completions.create(
    model="gpt-5-6-terra",
    messages=[{"role": "user", "content": "Resume esta factura en tres frases."}],
)
print(resp.choices[0].message.content)
print(resp.usage)  # tokens de entrada, de salida y de razonamiento: la base del cobro

The key (sk-kn-…) is created in the dashboard after registration; you can also set a monthly spending limit per key there. The steps are in the quickstart.

When it makes sense to go directly to OpenAI

Kunavo is a shared-capacity gateway: there is no dedicated quota for your account and no contractual SLA. If you need guaranteed quota or an SLA — a product with committed availability for your customers — buy directly from OpenAI. To redo the calculations with your figures, the OpenAI API pricing calculator includes cached input and reasoning tokens. The equivalent for Anthropic: Claude pricing.

Frequently asked questions