In this article
Choosing between OpenAI, Anthropic, and Google for your AI application is no longer just about capability. All three offer excellent models, and the cost difference between them, and between each provider's models, can be more than a hundredfold. Knowing how to compare prices is what separates a profitable product from an out-of-control budget.
This article compares the prices of the main AI APIs, with current values as of October 2026. As these prices change frequently, always check the official page of each provider before making a final decision.
How to read the prices
All providers charge per million tokens, with separate values for input (what you send) and output (what the model generates). The usual notation is input / output. A value of $3 / $15 means three dollars per million input tokens and fifteen per million output tokens.
Remember that output almost always costs five times more than input, and that one million tokens is equivalent to about 750,000 words.
Comparison table (October 2026)
Values in dollars per million tokens, input / output, before caching or batch discounts.
| Provider | Model | Input | Output | Positioning |
|---|---|---|---|---|
| Anthropic | Claude Fable 5.1 | $10 | $50 | High-end |
| Anthropic | Claude Opus 5.5 | $4 | $20 | High-end |
| Anthropic | Claude Sonnet 5.5 | $2 | $10 | Balance for production |
| Anthropic | Claude Haiku 5.5 | $0.10 | $0.50 | Volume and low cost (up to 100K input tokens) |
| OpenAI | GPT-6 astra | $10 | $50 | High-end |
| OpenAI | GPT-6.1 sol | $2 | $10 | Balance for production |
| OpenAI | GPT-5.4 | $2.50 | $15 | Workhorse |
| OpenAI | GPT-5.4 nano | $0.20 | $1.25 | Ultra-economical |
| OpenAI | GPT-6 luna | $0.10 | $0.50 | Ultra-economical |
| Gemini 4 Argon | $2 | $10 | High-end (promo until 30 Nov 2026; $4 / $20 from 1 Dec) | |
| Gemini 3.1 Pro | $2 | $12 | High-end (preview) | |
| Gemini 3.8 Flash | $0.75 | $3.75 | Balance (promo until 31 Dec 2026; $1.50 / $7.50 from 1 Jan 2027) | |
| Gemini 3.5 Flash-Lite | $0.30 | $2.50 | Economical | |
| Gemini 2.5 Flash-Lite | $0.10 | $0.40 | Cheapest |
What this table shows
At the high end, the ceiling moved up. Claude Fable 5.1 and GPT-6 astra, both at $10 / $50, are now the most expensive models in the table. Just below, Opus 5.5 sits at $4 / $20. The new Gemini 4 Argon costs $2 / $10 until 30 November and, until then, is the cheapest high-end model of the three providers. On 1 December it moves to $4 / $20.
In economical models, Google leads. Gemini 2.5 Flash-Lite, at $0.10 / $0.40, is still the cheapest in the table, with GPT-6 luna and the new Claude Haiku 5.5 (both at $0.10 / $0.50) close behind. For Haiku 5.5, that price applies to requests up to 100K input tokens; above that it rises to $0.50 / $2.50. For simple, high-volume tasks, it's hard to beat on price. In the mid tier, Gemini 3.8 Flash costs $0.75 / $3.75 until 31 December and moves to $1.50 / $7.50 on 1 January 2027.
Watch the promotional prices. Gemini 4 Argon and Gemini 3.8 Flash have launch prices with an end date (30 November and 31 December 2026). If your project will run past those dates, do the maths with the next price, which is double.
The biggest factor is not the provider, it's the model. Notice that within OpenAI itself, GPT-6 astra costs a hundred times more than GPT-6 luna. Choosing the right model within a provider has more impact on cost than choosing between providers.
The discounts that change everything
List prices are just the starting point. All three providers offer mechanisms that drastically reduce effective cost:
- Batch: asynchronous processing with 50 percent discount, across all providers.
- Prompt caching: reuse of repeated context with 90 percent or more savings on reads.
By combining both, it's possible to reduce the cost of suitable workloads by more than 90 percent compared to the list price.
How to choose
The right question is not which is the cheapest overall, but which is the cheapest for your task, at the quality you need. A practical approach:
- Define the task and the minimum acceptable quality. Not everything needs the top-tier model.
- Test economical models first. Often a cheap model solves simple tasks like classification, extraction, or support responses.
- Reserve high-end models for complex reasoning, difficult code, or high-value work.
- Apply caching and batch where the usage pattern allows.
Conclusion
In 2026, price is no longer a barrier to using quality AI. Token costs have fallen sharply, and there are options for all budgets. The smart decision is not to commit to one provider, but to choose the right model for each task and apply the available savings mechanisms.
As values change frequently, treat this table as a snapshot from October 2026 and always confirm official prices before finalizing an architecture.
From the token bill to the project cost
Tokens are the small part of an AI project's cost. Most of it sits in integrating with the systems you already have, preparing the data and changing how the team works. To do the full sum, see how much it costs to implement AI in a company, how to measure AI ROI and the automation ROI calculator.
Sources
Prices checked on 27 October 2026 on the providers' official pages:
To run the numbers for your own case, use the AI token cost calculator.
Writes about AI consulting for operations and processes: how to map and simplify processes, apply AI where it has the most impact and measure the gains.