For developers · verified September 25, 2026

AI API pricing: 29 models, per million tokens.

Input and output prices for OpenAI, Anthropic, Google, xAI, DeepSeek and Mistral models, sorted cheapest first. Estimate your monthly bill with the calculator below.

Provider

Sort

Model	Provider	Input /1M	Output /1M	Context	Notes
GPT-6 Luna	OpenAI	$0.10	$0.50	1.05M	Cheapest GPT-6 model.
DeepSeek V4.1-Flash	DeepSeek AI	$0.15	$0.60	1M	Off-peak price. Peak hours (weekdays 01–04 & 06–10 UTC): $0.30 / $1.20. Replaced V4-Flash.
Mistral Small 4	Mistral AI	$0.15	$0.60	256K
GPT-5.6 Luna	OpenAI	$0.20	$1.20	1.05M
GPT-5.4 nano	OpenAI	$0.20	$1.25	400K
Gemini 3.1 Flash-Lite	Google	$0.25	$1.50	1M
Gemini 3.5 Flash-Lite	Google	$0.30	$2.50	1M
Mistral Large 3	Mistral AI	$0.50	$1.50	256K
DeepSeek V4-Pro	DeepSeek AI	$0.66	$1.98	1M	Off-peak price. Peak: $1.32 / $3.96. Planned Sept 14 retirement was reversed.
Gemini 3.8 Flash Promo price	Google	$0.75	$3.75	1M	Launch price until Dec 31, 2026; then $1.50 / $7.50.
Gemini 3.7 Flash Promo price	Google	$0.75	$3.75	1M	Launch price until Dec 31, 2026; then $1.50 / $7.50.
Gemini 3.6 Flash Promo price	Google	$0.75	$3.75	1M	Launch price until Dec 31, 2026; then $1.50 / $7.50.
Claude Haiku 4.5	Anthropic	$1	$5	200K
Grok 4.3	xAI	$1.25	$2.50	1M
Mistral Medium 3.5	Mistral AI	$1.50	$7.50	256K
Gemini 3.5 Flash Legacy	Google	$1.50	$9	1M	Now labelled legacy; newer Flash models are cheaper.
Grok 4.7	xAI	$2	$6	500K	Above 200K tokens: $4 / $12.
Grok 4.5	xAI	$2	$6	500K
GPT-6 Sol	OpenAI	$2	$10	1.05M	Balanced GPT-6 tier.
Claude Sonnet 5	Anthropic	$2	$10	1M	Planned increase to $3 / $15 was cancelled; $2 / $10 is permanent.
GPT-5.6 Terra	OpenAI	$2	$12	1.05M
Gemini 3.1 Pro Preview	Google	$2	$12	1M	Preview. Above 200K tokens: $4 / $18.
GPT-5.6 Sol Promo price	OpenAI	$4	$20	1.05M	Promotional price until Nov 21, 2026; list price is $5 / $30.
Claude Opus 5.5	Anthropic	$4	$20	1M	New Opus; cheaper than Opus 5.
Claude Opus 5	Anthropic	$5	$25	1M
Claude Opus 4.8	Anthropic	$5	$25	1M
GPT-6 Astra	OpenAI	$10	$50	1.05M	OpenAI's most capable model. Long context above 272K costs more.
Claude Fable 5.1	Anthropic	$10	$50	1M	Anthropic's top generally available tier.
GPT-5.5 pro	OpenAI	$30	$180	1.05M	Most expensive model in this table.

Blended = (3 × input + output) ÷ 4, a typical chat mix. Off-peak prices shown for DeepSeek. Promotional prices are marked.

API cost calculator

Requests per day Input tokens per request Output tokens per request Days per month

Model	Provider	Per month	Per year

For reference, 1M input + 1M output tokens costs: GPT-6 Luna ≈ $0.60 · DeepSeek V4.1-Flash ≈ $0.75 · Gemini 3.8 Flash ≈ $4.50 · GPT-6 Sol ≈ $12.00 · Claude Sonnet 5 ≈ $12.00 · Claude Opus 5.5 ≈ $24.00 · GPT-6 Astra ≈ $60.00.

API pricing questions

How do API prices work?

You pay per million tokens (roughly 750,000 English words). Input (what you send) and output (what the model writes) are priced separately; output usually costs 3–6× more.

What's the cheapest capable API model right now?

By input price, GPT-6 Luna at $0.10 in / $0.50 out. GPT-6 Luna ($0.10 / $0.50) and DeepSeek V4.1-Flash ($0.15 / $0.60 off-peak) are the budget standouts.

Which API prices are temporary?

GPT-5.6 Sol is on a promotion until Nov 21, 2026 (list $5 / $30). Gemini 3.6, 3.7 and 3.8 Flash are $0.75 / $3.75 until Dec 31, 2026, then $1.50 / $7.50. DeepSeek charges 2× in weekday peak hours.

Do I need the API or a subscription?

If you chat with AI yourself, a subscription is almost always cheaper. The API is for developers building products or automations.

Sources

Checked against the official pages below on September 25, 2026. Prices are USD before local tax. See how we verify prices and the change log.