Models
186 models with live API prices, independent benchmarks and real usage. Filter by what you're building and see your monthly cost.
Prices tracked from the official pages of
Compare AI models
Pick a use case, set your volume, and compare quality against what you'll actually pay.
Build software
Create media
Run in production
Every token-priced card shows its cost for this workload.
Based on Artificial Analysis scores and your workload estimate
Top performer
Highest Intelligence score among the models in view.
Best value
Most Intelligence per dollar — about 40× cheaper than the top performer.
Budget pick
Cheapest model that still scores above the median on Intelligence.
174 models · sorted by best fit
Intelligence
58.4
top 1%
Web design
1325 Elo
#5 of 172
UI components
1378 Elo
#3 of 157
Input
$4
Cached input
$0.20
Output
$20
per 1M tokens · Standard
686× the cheapest in view
Intelligence
56.9
top 4%
Coding
77.1
top 5%
Agentic
51
top 9%
Input
$10
Cached input
$1
Output
$50
per 1M tokens · Standard
1730× the cheapest in view
Intelligence
54.5
top 8%
Web design
1314 Elo
#9 of 172
UI components
1342 Elo
#8 of 157
Input
$2
Cached input
$0.10
Output
$10
per 1M tokens · Standard
343× the cheapest in view
Intelligence
54.2
top 10%
Web design
1321 Elo
#7 of 172
UI components
1375 Elo
#4 of 157
Input
$2
Cached input
$0.10
Output
$10
per 1M tokens · Standard
343× the cheapest in view
Intelligence
51
top 16%
Input
$1
Cached input
$0.05
Output
$5
per 1M tokens · Standard
172× the cheapest in view
Our flagship reasoning model, built for complex software engineering, deep mathematical and scientific problem solving, autonomous multi-step agents, and long-decode generation up to 1M tokens.
Intelligence
50.4
top 18%
Web design
1315 Elo
#8 of 172
UI components
1332 Elo
#14 of 157
Input
$2
Cached input
$0.10
Output
$10
per 1M tokens · Standard
343× the cheapest in view
Intelligence
49.7
top 23%
Coding
81.6
top 1%
Agentic
57.9
top 1%
Input
$10
Cached input
$0.25
Output
$50
per 1M tokens · Standard
1708× the cheapest in view
Intelligence
36.2
top 37%
Web design
1262 Elo
#39 of 172
UI components
1278 Elo
#27 of 157
Input
$0.10
Cached input
$0.01
Output
$0.50
per 1M tokens · Standard · for prompts up to 100,000 tokens
17× the cheapest in view
Intelligence
35
top 44%
Web design
1248 Elo
#47 of 172
UI components
1266 Elo
#36 of 157
Input
$2
Cached input
$0.50
Output
$6
per 1M tokens · Standard · < 200k prompt tokens
275× the cheapest in view
Our most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows.
Intelligence
30.5
top 49%
Coding
76.3
top 9%
Agentic
40.2
top 28%
Input
$0.75
Cached input
$0.075
Output
$3.75
per 1M tokens · Standard
130× the cheapest in view
Intelligence
27.8
top 62%
Input
$0.10
Cached input
$0.01
Output
$0.50
per 1M tokens · Standard
17× the cheapest in view
Intelligence
23.1
Coding
68.8
Agentic
41.3
Input
$1.32
Cached input
$0.044
Output
$3.96
per 1M tokens · Standard
173× the cheapest in view
Intelligence
20.1
top 77%
Coding
68.1
top 28%
Agentic
45.8
top 19%
Input
$0.80
Cached input
—
Output
$4
per 1M tokens · Standard
160× the cheapest in view
A cost-efficient model, optimized for high-volume agentic tasks, translation, and simple data processing.
Intelligence
11.7
top 92%
Coding
49.3
top 51%
Agentic
14.3
top 64%
Input
$0.30
Cached input
$0.03
Output
$2.50
per 1M tokens · Standard
72× the cheapest in view
Intelligence
4.3
top 100%
Coding
26.6
top 73%
Agentic
1.2
top 85%
Input
$0.15
Cached input
—
Output
$0.60
per 1M tokens · Standard
27× the cheapest in view
Input
$0.02
Cached input
—
Output
$0
per 1M tokens · Standard
2.0× the cheapest in view
Coding
20.1
top 83%
Agentic
1
top 90%
Web design
1162 Elo
#103 of 172
Input
$0.50
Cached input
—
Output
$1.50
per 1M tokens · Standard
80× the cheapest in view
Input
$0.70
Cached input
$0.07
Output
$2.10
per 1M tokens · Standard
93× the cheapest in view
Gemini Robotics ER 2, short for Gemini Robotics Embodied Reasoning 2, is a vision-language model endpoint that enables robots to understand their environments precisely, supporting agentic orchestration of robots, video progress understanding, multi-robot collaboration, and advanced spatial reasoning.
Input
$1
Cached input
$0.10
Output
$5
per 1M tokens · Standard
173× the cheapest in view
Input
$1.40
Cached input
$0.14
Output
$4.40
per 1M tokens · Standard
190× the cheapest in view
Gemini Robotics ER 2 Streaming is a vision-language model endpoint for robotics optimized for real-time text streaming using the Live API. It accepts text, image, video, and audio input and supports bidirectional streaming with function calling.
Input
$1
Cached input
—
Output
$5
per 1M tokens · Standard
200× the cheapest in view
Coding
72.4
top 17%
Agentic
41.2
top 26%
Web design
1280 Elo
#27 of 172
Input
$2
Cached input
$0.30
Output
$6
per 1M tokens · Standard · < 200k prompt tokens
269× the cheapest in view
Coding
76.8
top 6%
Agentic
53
top 6%
Web design
1288 Elo
#23 of 172
Input
$2
Cached input
$0.50
Output
$6
per 1M tokens · Standard · < 200k prompt tokens
275× the cheapest in view
A Pro model which excels at coding and complex reasoning tasks.
Coding
33.3
top 68%
Agentic
1.6
top 81%
Web design
1165 Elo
#100 of 172
Input
$1.25
Cached input
$0.125
Output
$10
per 1M tokens · Standard
291× the cheapest in view
Not sure where to begin?
What you want to build, how much you code and your budget: you'll get the exact tools and the model to use.
Step 1 of 3
AI 101
No jargon. Read this once and every page on AgentCost will make sense.
The “brain”. You send it text (or images, audio) and it sends back an answer. GPT, Claude, Gemini and DeepSeek are model families.
“Summarize this email” → a summary.
A model plus tools and a loop: it plans, takes actions (edit files, browse, call APIs), checks results and keeps going until the task is done.
“Add a login page” → it writes code, runs it, fixes errors.
How your app talks to a model. You pay per use instead of a monthly seat, which is why your volume decides the bill.
Your support page sends each question to the model's API.
The unit models are billed in — roughly ¾ of a word. Prices are “per 1M tokens”, with input (what you send) and output (what it writes) priced separately.
A 750-word article ≈ 1,000 tokens.
How much the model can read at once — its short-term memory. Bigger windows fit whole codebases or long documents in one request.
1M tokens ≈ several novels at once.
A standardized test that scores models on the same tasks, so you can compare quality fairly instead of trusting marketing.
Coding Index: how well a model solves real coding tasks.
You want an answer, text or a classification
→ A model
Call its API, or chat with it
You want a task done end-to-end
→ An agent
Use a ready agent or build one with a framework
You don't code and want a working app
→ An app-builder agent
Describe it, iterate by chatting
You want to cut cost
→ A smaller model
Use the “Best value” sort and test on real data
Usage trends
Real token volume across apps routed through OpenRouter (2026-10-29 – 2026-10-29). Not a survey — what production traffic is choosing right now.
Tokens processed in the latest ranking window.
The tools developers build with every day.
Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.
Claude Code is Anthropic's agentic coding tool that reads your entire codebase, plans and executes changes across files, runs tests, and iterates on failures, all from natural language prompts.
There are many coding agents, but this one is yours.
A coding agent that helps you build and ship with AI
Kilo Code is an open-source AI coding agent that works across VS Code, JetBrains, and CLI to help developers ship code faster with agentic workflows.
OpenClaw is an open-source AI agent that connects to your messaging apps and takes real actions on your behalf, from running commands and browsing the web to managing files and sending emails.
Methodology
Straight from each provider's official pricing page, re-checked every few hours. Every model's “All prices” panel links the exact page it came from. If a page changes and our reader breaks, we keep showing the last verified prices and mark the provider below.
We don't run benchmarks and never make up numbers. Intelligence, Coding and Agentic scores are Artificial Analysis's independent indexes; web and UI design scores are Design Arena's crowd-voted Elo ratings; “most used” ranks are real token volume from OpenRouter. Each is linked below.
From the models currently in view: the top performer has the highest score on the benchmark that fits your use case; best value has the most score per dollar among above-median models; the budget pick is the cheapest model that still scores above the median. Change filters or your workload and the picks update.
A model's position in its own provider's lineup, read from the model family (Opus/Sonnet/Haiku, Pro/Flash/Lite, -mini…) and its price. It's a quick orientation aid — for actual quality, use the benchmark scores.
Your monthly input and output tokens × the model's per-million-token prices, with the cached share of input charged at the cache-read rate. It excludes tool fees, search grounding, images and taxes — always confirm on the provider's page before committing budget.
No. It's independent and free to use. No provider pays for placement or rankings.
Independent Intelligence, Coding and Agentic indexes
Crowd-voted Elo for websites, UI components and other design tasks
Real token usage across apps routed through OpenRouter
Community
Every testimonial is verified by hand before it's published.
Reach the developer
Spotted a wrong price, want a provider added, or have an idea that would make this more useful for your team? Partnerships and collaborations are welcome too.
Your story could be first here.
AgentCost just launched. If it helped you pick a model or cut your AI bill, tell other builders — your words will be the first thing the next visitor reads.