How is AI model pricing measured?

Per million tokens, a token being a piece of a word. Input (what you send) and output (what the model writes) are priced separately.

What is the blended price?

It weighs input and output 3 to 1, as Artificial Analysis does, so models can be compared with one number.

What is the cheapest AI model?

The one at the top of this board, which lists every model cheapest first. Makers change their prices often, so check the table for today's cheapest.

Why does output cost more than input?

Writing each token takes a model more work than reading one, so most makers charge more for output.

How many words is a million tokens?

About 750,000 English words, as a rule of thumb. Other languages often take more tokens per word.

What does the chart show?

Each dot is a model, placed by its intelligence score and its price. The line joins the models no cheaper model beats, so anything under it costs more for less.

How often is this board updated?

Every week, from OpenRouter's public model list. The date under the table shows when the prices was last checked.

Can I download the data?

Yes: every board has a CSV of its models, next to its chart.

What does open weights mean?

The maker publishes the model's weights, so anyone can download the model and run it on their own hardware.

What is a context window?

It's the most text a model can work with at once, counted in tokens. Your prompt and the model's reply both count toward it.

Why do some models have no score?

Not every model on OpenRouter has been tested on every measure yet. A board lists only the models with a result on its own measure.

How do I use these models?

Through each maker's own app or API, or through OpenRouter, which gives you most of them with one account.

AI model pricing

Compare AI model pricing for 365 AI models, cheapest first, with each model's price per million tokens and its intelligence score where it has one.

Updated Oct 1, 2026

Rank	Model
1	Mistral NemoOpen weightsMistral	–	$0.02	$0.03	131K
2	Ling 3.0 Flash VLOpen weightsinclusionAI	24.6	$0.02	$0.06	262K
3	Ling 3.0 FlashOpen weightsinclusionAI	20.1	$0.02	$0.06	262K
4	gpt-oss-20bOpen weightsOpenAI	9.0	$0.02	$0.09	131K
5	Granite 4.0 MicroOpen weightsIBM	–	$0.02	$0.11	131K
6	LLlama 3 8B LunarisOpen weightsSao10K	–	$0.04	$0.05	8K
7	NNex-N2.5-MiniOpen weightsNex AGI	–	$0.03	$0.10	262K
8	DeepSeek V4 Flash 0423Open weightsDeepSeek	24.4	$0.04	$0.08	1M
9	Qwen3.7 FlashQwen	–	$0.03	$0.13	1M
10	Llama 3.1 8B InstructOpen weightsMeta	–	$0.05	$0.08	131K
11	Mistral Small 3Open weightsMistral	–	$0.05	$0.08	33K
12	Schematron V2 TurboOpen weightsInference.net	–	$0.03	$0.15	128K
13	Nova Micro 1.0Amazon	–	$0.04	$0.14	128K
14	Gemma 3 4BOpen weightsGoogle	–	$0.05	$0.10	131K
15	Command R7B (12-2024)Cohere	–	$0.04	$0.15	128K
16	Mercury 2.5Inception	12.3	$0.04	$0.15	260K
17	gpt-oss-120bOpen weightsOpenAI	11.6	$0.04	$0.17	131K
18	Llama 3.2 1B InstructOpen weightsMeta	–	$0.03	$0.20	60K
19	Gemma 3 12BOpen weightsGoogle	3.8	$0.05	$0.15	131K
20	Laguna XS 2.1Open weightsPoolside	–	$0.06	$0.12	262K
21	Hy-MT2-1.8BOpen weightsTencent	–	$0.04	$0.18	8K
22	Qwen3 30B A3B Instruct 2507Open weightsQwen	–	$0.05	$0.19	262K
23	Nemotron 3.5 LightningOpen weightsNVIDIA	12.9	$0.06	$0.17	262K
24	MMythoMax 13BOpen weightsgryphe	–	$0.08	$0.11	8K
25	Phi 4Open weightsMicrosoft	5.9	$0.07	$0.14	16K

Source: Artificial Analysis's indexes and OpenRouter's prices, in US dollars per million tokens, checked Oct 1, 2026

Data by

Price per million tokens · OpenRouter

Score against price

Oct 1, 2026

Claude Opus 5.5 · $8Claude Sonnet 5.5 · $4Claude Fable 5.1 · $20Gemini 4 Argon · $4Claude Fable 5 · $20Muse Spark 1.3 · $2MiMo-V2.6-Pro · $0.54DeepSeek V4.1 Flash · $0.15Hy3 · $0.14

Source: Artificial Analysis's indexes and OpenRouter's prices, in US dollars per million tokens, checked Oct 1, 2026

Data by

How to read this chart

Each dot is a model, placed by its price and its score. The models on the line are the value leaders: no cheaper model scores higher. Look toward the upper left for the best score for what you pay.

Comparison details

Prices are OpenRouter's list prices in US dollars per million tokens, blended 3 parts input to 1 part output, as Artificial Analysis does. Scores are Artificial Analysis's, checked Oct 1, 2026. Speed, latency and each provider's own price aren't part of the score.

Every leaderboard

Each board has its own page with every model, a chart and a CSV.

LLM leaderboard

Claude Opus 5.5 leads, from Anthropic

Best AI model for coding

Claude Fable 5.1 leads, from Anthropic

Best AI model for agents

Claude Fable 5.1 leads, from Anthropic

Best AI model for writing

Best AI model for website design

Muse Spark 1.3 leads, from Meta

Best AI model for game development

Claude Opus 5.5 leads, from Anthropic

Best AI model for UI design

Claude Opus 5.5 leads, from Anthropic

Best AI model for data visualization

Claude Opus 5.5 leads, from Anthropic

Best AI model for SVG graphics

Claude Opus 5.5 leads, from Anthropic

Best AI model for 3D scenes

Claude Opus 5.5 leads, from Anthropic

Best AI model for building mobile apps

Claude Opus 5 leads, from Anthropic

Best AI model for building full-stack apps

Claude Fable 5.1 leads, from Anthropic

Best AI image model

GPT Image 2.5 Sunburst (max) leads, from OpenAI

Best AI image editing model

GPT Image 2.5 Sunburst (max) leads, from OpenAI

Best AI video model

Wan 3.0 leads, from Alibaba

Best image to video AI model

Gemini Omni Flash leads, from Google

Best AI voice model

Eleven v4 leads, from ElevenLabs