Reference / AI Model Pricing

AI model pricing per million tokens

Q: What is the difference between input and output tokens?

Input tokens are what you send to the model. Output tokens cost 3 to 5 times more than input tokens.

Q: What is the batch API discount?

OpenAI and Anthropic offer 50% off for async batch requests. Document processing, classification and content generation typically qualify.

Every major LLM priced by input and output tokens. Standard API rates with no volume discounts applied unless noted. Prices monitored continuously.

OpenAI models

Model	Input / 1M tokens	Output / 1M tokens	Context	Batch discount
GPT-5	$1.25	$10.00	400k	Yes, 50% off
GPT-4.1	$2.00	$8.00	1M	Yes, 50% off
GPT-4.1 mini	$0.40	$1.60	1M	Yes, 50% off
GPT-4.1 nano	$0.10	$0.40	1M	Yes, 50% off
o4-mini	$1.10	$4.40	200k	Yes, 50% off

Anthropic models

Model	Input / 1M tokens	Output / 1M tokens	Context	Batch discount
Claude Opus 4.6	$5.00	$25.00	1M	Yes, 50% off
Claude Sonnet 4.6	$3.00	$15.00	1M	Yes, 50% off
Claude Haiku 4.5	$1.00	$5.00	200k	Yes, 50% off

Google models

Model	Input / 1M tokens	Output / 1M tokens	Context	Batch discount
Gemini 2.5 Pro	$1.25	$10.00	1M	No
Gemini 2.5 Flash	$0.30	$2.50	1M	No
Gemini 2.0 Flash	$0.10	$0.40	1M	No

Other models

Model	Input / 1M tokens	Output / 1M tokens	Notes
Llama 4 Scout (Meta)	$0.15	$0.50	10M context. Open source.
Mistral Large 3	$0.50	$1.50	EU data residency option.
DeepSeek V3	$0.28	$0.42	Cheapest available. Data residency risk.

Common questions about AI model pricing

What is the cheapest AI API right now?

Gemini 2.0 Flash and GPT-4.1 nano are both $0.10 per million input tokens and $0.40 per million output tokens, the cheapest capable models from major providers. DeepSeek V3 is cheaper but carries data residency risks.

What is the difference between input and output tokens?

Input tokens are what you send to the model including your prompt, context and documents. Output tokens are what the model generates back. Output tokens cost 3 to 5 times more than input tokens across most APIs.

What is the batch API discount?

OpenAI and Anthropic both offer 50% off for asynchronous batch requests that do not need an instant response. Document processing, classification and content generation typically qualify.

How much does GPT-4.1 cost per million tokens?

GPT-4.1 costs $2.00 per million input tokens and $8.00 per million output tokens at standard API rates. With the 50% batch discount that drops to $1.00 input and $4.00 output per million tokens.

Prices monitored continuously. All figures in USD at standard API rates.