AI model API cost calculator

Enter your workload: the calculator applies the list prices published by each vendor and ranks every model from cheapest to most expensive.

Prices verified Oct 20, 2026

Model

Tasks per month

Input tokens per task

Output tokens per task

Prefilled values = an example. 200 tasks of 50,000 input and 5,000 output tokens: an arbitrary order of magnitude, not a measurement. Replace them with your own numbers (your provider's console shows the tokens used).

Estimated monthly cost

Claude Sonnet 5.5: $30.00 per month · cheapest rank 19 of 26

Monthly cost = tasks × (input tokens × input price + output tokens × output price) ÷ 1,000,000.

Every model, cheapest first

Every model, cheapest first
Rank	Model	Vendor	Input price ($/M tokens)	Output price ($/M tokens)	Monthly cost	Pricing conditions	Sources
1	Claude Haiku 5.5	Anthropic	$0.1	$0.5	$1.50	Tiered or temporary price: base price used. prompts ≤ 100K tokens; $0.50/$2.50 above	price — Claude Haiku 5.5
2	GPT-6 Luna	OpenAI	$0.1	$0.5	$1.50	—	price — GPT-6 Luna
3	Qwen3.8-Flash	Alibaba (Qwen)	$0.15	$0.47	$1.97	international (Singapore) price	price — Qwen3.8-Flash
4	GLM-5.3-Flash	Z.ai	$0.15	$0.5	$2.00	—	price — GLM-5.3-Flash
5	MiniMax M3.1 Flash	MiniMax	$0.15	$0.6	$2.10	flat rate; $0.015 cached input	price — MiniMax M3.1 Flash
6	Mistral Small 4	Mistral AI	$0.15	$0.6	$2.10	—	price — Mistral Small 4
7	Codestral 25.08	Mistral AI	$0.3	$0.9	$3.90	code completion (FIM)	price — Codestral 25.08
8	DeepSeek V4.1-Flash-1017	DeepSeek	$0.3	$1.2	$4.20	Tiered or temporary price: base price used. peak hours, cache miss; half price off-peak	price — DeepSeek V4.1-Flash-1017
9	MiniMax M3	MiniMax	$0.3	$1.2	$4.20	Tiered or temporary price: base price used. input ≤ 512K tokens; $0.60/$2.40 above	price — MiniMax M3
10	Gemini 3.5 Flash-Lite	Google	$0.3	$2.5	$5.50	—	price — Gemini 3.5 Flash-Lite
11	Qwen3.8-27B	Alibaba (Qwen)	$0.5	$3	$8.00	international (Singapore) price	price — Qwen3.8-27B
12	Gemini 3.8 Flash	Google	$0.75	$3.75	$11.25	Tiered or temporary price: base price used. through Dec 31, 2026, then $1.50/$7.50	price — Gemini 3.8 Flash
13	DeepSeek V4-Pro-0813	DeepSeek	$1.32	$3.96	$17.16	Tiered or temporary price: base price used. peak hours, cache miss; half price off-peak	price — DeepSeek V4-Pro-0813
14	Mistral Large 4	Mistral AI	$1.36	$4.18	$17.78	public preview	price — Mistral Large 4
15	GLM-5.3	Z.ai	$1.4	$4.4	$18.40	—	price — GLM-5.3
16	Mistral Medium 3.5	Mistral AI	$1.5	$7.5	$22.50	—	price — Mistral Medium 3.5
17	Grok 4.7	xAI	$2	$6	$26.00	—	price — Grok 4.7
18	Qwen3.8-Max	Alibaba (Qwen)	$2	$6	$26.00	international (Singapore) price	price — Qwen3.8-Max
19	Claude Sonnet 5.5 selected	Anthropic	$2	$10	$30.00	—	price — Claude Sonnet 5.5
20	Gemini 4 Argon	Google	$2	$10	$30.00	Tiered or temporary price: base price used. launch price until Nov 30, 2026, then $4/$20	price — Gemini 4 Argon
21	GPT-6.1 Sol	OpenAI	$2	$10	$30.00	Tiered or temporary price: base price used. above 272K input tokens: 2x input, 1.5x output	price — GPT-6.1 Sol
22	GPT-5.3-Codex	OpenAI	$1.75	$14	$31.50	specialized Codex model	price — GPT-5.3-Codex
23	Kimi K3	Moonshot AI	$3	$15	$45.00	cache-miss input; $0.30 cache hit	price — Kimi K3
24	Claude Opus 5.5	Anthropic	$4	$20	$60.00	—	price — Claude Opus 5.5
25	Claude Fable 5.1	Anthropic	$10	$50	$150.00	—	price — Claude Fable 5.1
26	GPT-6 Astra	OpenAI	$10	$50	$150.00	Tiered or temporary price: base price used. above 272K input tokens: 2x input, 1.5x output	price — GPT-6 Astra

Indicative estimate in US dollars, excluding tax, caching, batch processing, off-peak rates and subscriptions. Cache discounts are not computed: our data only lists them as notes for a few models. An agent task often chains several requests: tiered-pricing thresholds apply per request, not per task. · Verified Oct 20, 2026

Price sources

Model comparison → Price / performance chart → Benchmarks →