OpenAI API Pricing (2026): GPT-6 Astra, GPT-6.1 Sol, GPT-6 Luna and the Full Per-Token Table

The Current OpenAI API Price Table

The table below is the current OpenAI API lineup as listed on the official pricing page, in USD per million tokens at the Standard tier, short-context rate. Input is the cost to send tokens to the model. Cached input is the discounted rate for a repeated prompt prefix. Cache writes bill the first time a prefix is stored. Output is the cost of generated tokens. All figures verified October 15, 2026.

The spread across the lineup is wide. gpt-6-luna output at $0.50/1M versus gpt-6-astra output at $50.00/1M is a 100x difference. Matching the model to the task beats defaulting every call to the flagship.

How OpenAI Token Pricing Works

OpenAI charges per token, not per request. A token is roughly 4 characters of English text, so 1,000 tokens is about 750 words. Every API call has an input token count (your prompt plus conversation history plus any tool schemas) and an output token count (the model's response). You pay for both at different rates.

Output costs 5x input on all three GPT-6 models. gpt-6-astra charges $10.00/1M input and $50.00/1M output. gpt-6.1-sol charges $2.00/1M input and $10.00/1M output. This asymmetry means a read-heavy workload (summarizing a large document, answering questions about code) is far cheaper than a generation-heavy workload (writing new files, producing long completions) at the same total token count.

Two more dimensions multiply on top of the base rate: the processing tier (Batch halves rates, Fast mode doubles them) and context length (long-context requests bill a higher second rate). Both are covered below.

Processing Tiers: Batch, Flex, and Fast Mode

OpenAI sells the same models at five speeds. Batch runs asynchronous jobs at 50% of Standard. Flex bills the same 50% for latency-tolerant online traffic. Fast mode, renamed from Priority processing on July 30, 2026, costs 2x Standard. Ultrafast costs 6x Standard and is listed for gpt-6-astra and gpt-6.1-sol.

On gpt-6-astra the same tiers run $5.00/$25.00 (Batch and Flex), $10.00/$50.00 (Standard), $20.00/$100.00 (Fast), and $60.00/$300.00 (Ultrafast). Regional processing (data residency) and FedRAMP endpoints add a 10% uplift for models released on or after March 5, 2026.

Long-Context Pricing

Every GPT-6 row carries two rates: a short-context rate and a higher long-context rate. Input and cached input double. Output rises 1.5x.

Retrieval, compaction, and summarization that keep a request in the short-context band avoid the 2x input rate. For agents that re-send a long history every turn, that saving compounds across the session.

What Happened to GPT-4o and GPT-5 Pricing

OpenAI's flagship pricing tables now show only the GPT-6 family. The GPT-5.6 series moved: gpt-5.6-sol is listed under Cyber models at a promotional $4.00/$20.00, and gpt-5.6-terra and gpt-5.6-luna no longer appear in the main tables. GPT-4o survives in the audio section as gpt-4o-transcribe. Older chat models are in OpenAI's All models list, so check it before budgeting a legacy model.

For capability planning, gpt-6-astra replaces the GPT-5.6 flagship slot, gpt-6.1-sol covers general work, and gpt-6-luna covers high-volume classification and extraction at $0.10/1M input.

Cached Input Discount

Cached input is the largest discount OpenAI exposes directly in the price table. When the same prompt prefix repeats across calls, OpenAI bills the cached portion at a fraction of the input rate. gpt-6.1-sol has the deepest discount at 95%: $0.10/1M cached versus $2.00/1M uncached. The first write of a prefix bills at 1.25x the input rate.

To capture the discount, put stable content first and variable content last. An agent that rebuilds its system prompt every turn, or interleaves changing context with fixed context, defeats the cache and pays the full input rate plus repeated cache writes.

Worked Example: A Coding Session

Take a coding session of 100 API calls totaling 300K input tokens and 100K output tokens, uncached, Standard tier.

The all-flagship session costs $8.00. The all-luna session costs $0.08, a 100x difference. Neither extreme fits a real workload: luna misses the hard prompts, and gpt-6-astra is wasted on the easy ones. Split the same 100 calls by difficulty:

The routed session costs $1.648 versus $8.00 all-flagship, a 79% reduction, with gpt-6-astra still on the hard 15%. Morph Router picks the model per prompt for $0.005 per request and returns the choice; your code makes the call.

Monthly Cost Scenarios

Monthly spend at Standard tier, uncached, assuming 1,000 input and 300 output tokens per request (light and medium) and 3,000 input and 1,000 output tokens per request (heavy agent traffic).

Batch or Flex halves every figure for work that tolerates latency: the heavy gpt-6.1-sol month drops from $4,800 to $2,400. Cached prefixes cut the input share further.

ChatGPT Plus vs the API

ChatGPT Plus at $20/month and pay-per-token API access are different products. Plus covers the ChatGPT app and Codex. The API is billed separately per token no matter what you subscribe to.

At gpt-6.1-sol rates, a request of 1,000 input and 300 output tokens costs $0.005 ($0.002 input + $0.003 output). $20 buys 4,000 such requests, about 133 per day for a month. On gpt-6-luna the same request costs $0.00025, so $20 covers 80,000 requests. Heavy agent use on gpt-6-astra passes $20 in a few sessions.

Calling the API

The OpenAI API is OpenAI-SDK-compatible. You set the model string, send messages, and receive a response with a usage block that reports exact input and output token counts. Use that usage block to verify your cost estimates against real traffic instead of guessing.

import OpenAI from "openai";

const openai = new OpenAI();

const response = await openai.chat.completions.create({
  model: "gpt-6.1-sol",        // $2.00/1M input, $10.00/1M output
  messages: [
    { role: "system", content: "You are a coding assistant." },
    { role: "user", content: "Rename fetchUsers to loadUsers across the file." },
  ],
});

// The usage block reports exact token counts for cost accounting
const { prompt_tokens, completion_tokens } = response.usage;
const inputCost = (prompt_tokens / 1_000_000) * 2.00;
const outputCost = (completion_tokens / 1_000_000) * 10.00;
console.log("Call cost: $" + (inputCost + outputCost).toFixed(6));

Morph serves its own fast open-weight coding models, including DeepSeek V4.1 Flash and GLM-5.3, on an OpenAI-compatible API at https://api.morphllm.com/v1. Point the OpenAI SDK's baseURL at it to send easy prompts to a cheaper model without changing your code.

import OpenAI from "openai";

// Point the OpenAI SDK at Morph's OpenAI-compatible endpoint
const client = new OpenAI({
  apiKey: process.env.MORPH_API_KEY,
  baseURL: "https://api.morphllm.com/v1",
});

const response = await client.chat.completions.create({
  model: "morph-dsv41flash",  // DeepSeek V4.1 Flash served by Morph
  messages: [
    { role: "user", content: "Add a docstring to the parseConfig function." },
  ],
});
// Same usage block, same cost accounting as the OpenAI call above.

How to Reduce OpenAI API Costs

Four levers move OpenAI API spend, in order of impact. The first three are structural and apply within OpenAI; the fourth works across providers.

The router lever is the one most teams skip because it sounds like rebuilding their stack. It is not. A router classifies each prompt and returns the cheapest model that clears the quality bar, so the easy work stops hitting the flagship. Morph's router classifies in ~430ms (~$0.001 per classification) and routes across many models through one OpenAI-compatible endpoint at api.morphllm.com, cutting spend 40-70% without touching the rest of your code. See how the LLM router works and the cost calculator.

Frequently Asked Questions

Sources

• OpenAI API pricing docs (developers.openai.com), fetched October 15, 2026

Related Resources

• What Is an LLM Router? Automatic Model Routing for Cost and Quality
• Anthropic API Pricing: Claude Per-Token Table
• Gemini API Pricing: Flash, Pro, and Flash-Lite
• DeepSeek API Pricing
• LLM Cost Calculator
• LLM Cost Optimization
• LLM Gateway: One Endpoint for Many Models
• Multi-Agent Model Routing
• LLM API Comparison
• Best AI Model for Coding