AI Model Atlas
Get started

Model & Pricing Atlas

Up-to-date model cards + pricing, normalized and source-linked.

API pricing Compare Subscriptions
How to read this page
API pricing is pay-as-you-go: you only pay for what you use. Prices are per 1M tokens (roughly 750,000 words, depending on language and formatting). Input is the text you send to the model; output is the text it generates for you. Some providers also offer cached input, which can reduce costs when you reuse the same text. Want a walkthrough? Follow our tutorial.

Featured pricing

USD · per 1M tokens · most expensive first Updated: 2026-03-12

Lower Higher
Open model
Anthropic • Claude Opus 4.6
$5.00 in
↗
Open model
Anthropic • Claude Opus 4.5
$5.00 in
↗
Open model
Anthropic • Claude Sonnet 4.6
$3.00 in
↗
Open model
xAI • Grok 4
$3.00 in
↗
Open model
OpenAI • GPT-4o
$2.50 in
↗
Open model
Google • Gemini 3 Pro Preview
$2.00 in
↗
Open model
OpenAI • GPT-5.4
$1.75 in
↗
Open model
OpenAI • GPT-5.2
$1.75 in
↗
Open model
Google • Gemini 2.5 Pro
$1.25 in
↗
Open model
Anthropic • Claude Haiku 4.5
$1.00 in
↗
Open model
Google • Gemini 3 Flash Preview
$0.60 in
↗
Open model
Google • Gemini 2.5 Flash
$0.30 in
↗
Open model
OpenAI • GPT-5 mini
$0.25 in
↗
Open model
xAI • Grok 4 Fast (Reasoning)
$0.20 in
↗
Open model
xAI • Grok 4.1 Fast (Reasoning)
$0.20 in
↗
Open model
xAI • Grok Code Fast 1
$0.20 in
↗
Bars are log-scaled within this view so very cheap models remain visible.

What is the best model?

See the live community leaderboard to compare performance across top models.

Open LM Arena →

Anthropic

Flagship models.

View all models →
Open model
Anthropic
Active
Claude Opus 4.6
200K ctx Reasoning Vision
Most capable Claude model. State-of-the-art reasoning, coding, and extended thinking for complex tasks.
Input
$5.00
Output
$25.00
Cached input
$0.50
Source ↗
per 1M tokens
Open model
Anthropic
Active
Claude Sonnet 4.6
200K ctx Reasoning Vision
Best balance of speed, intelligence, and cost. Ideal for production coding and analysis workloads.
Input
$3.00
Output
$15.00
Cached input
$0.30
Source ↗
per 1M tokens
Open model
Anthropic
Active
Claude Haiku 4.5
200K ctx Vision
Fastest Claude model. Optimized for high-throughput tasks like classification, extraction, and chat.
Input
$1.00
Output
$5.00
Cached input
$0.10
Source ↗
per 1M tokens
Open model
Anthropic
Active
Claude Opus 4.5
200K ctx Reasoning Vision
Previous flagship Claude model. Excellent at complex analysis, creative writing, and nuanced reasoning.
Input
$5.00
Output
$25.00
Cached input
$0.50
Source ↗
per 1M tokens

Google

Flagship models.

View all models →
Open model
Google
Preview
Gemini 3 Pro Preview
1.0M ctx Reasoning Vision Audio
Google's most capable model. 1M+ context, strong reasoning, and multimodal understanding.
Input
$2.00
Output
$12.00
Cached input
$0.40
Source ↗
per 1M tokens
Open model
Google
Preview
Gemini 3 Flash Preview
1.0M ctx Vision Audio
Fast, efficient Gemini model with 1M+ context window. Great for high-volume production tasks.
Input
$0.60
Output
$3.00
Cached input
$0.06
Source ↗
per 1M tokens
Open model
Google
Active
Gemini 2.5 Pro
1.0M ctx Reasoning Vision Audio
Advanced Gemini model with strong reasoning and 1M token context. Excellent for complex tasks.
Input
$1.25
Output
$10.00
Cached input
$0.13
Source ↗
per 1M tokens
Open model
Google
Active
Gemini 2.5 Flash
1.0M ctx Reasoning Vision Audio
Cost-effective Gemini model balancing speed and capability. Large context at low cost.
Input
$0.30
Output
$2.50
Cached input
$0.03
Source ↗
per 1M tokens

OpenAI

Flagship models.

View all models →
Open model
OpenAI
Preview
GPT-5.4
128K ctx Reasoning Vision
Most capable OpenAI model. Excels at complex reasoning, coding, and creative tasks with broad world knowledge.
Input
$1.75
Output
$14.00
Cached input
$0.18
Source ↗
per 1M tokens
Open model
OpenAI
Preview
GPT-5.2
128K ctx Reasoning Vision
High-capability model balancing cost and performance. Strong at analysis, writing, and multi-step reasoning.
Input
$1.75
Output
$14.00
Cached input
$0.18
Source ↗
per 1M tokens
Open model
OpenAI
Active
GPT-5 mini
128K ctx Vision
Fast, cost-efficient model for everyday tasks. Great price-to-performance for high-volume use cases.
Input
$0.25
Output
$2.00
Cached input
$0.03
Source ↗
per 1M tokens
Open model
OpenAI
Active
GPT-4o
128K ctx Vision Audio
Multimodal model with text, vision, and audio capabilities. Fast and versatile for production use.
Input
$2.50
Output
$10.00
Cached input
$1.25
Source ↗
per 1M tokens

xAI

Flagship models.

View all models →
Open model
xAI
Active
Grok 4
131.1K ctx Reasoning Vision
xAI's flagship model. Strong reasoning with real-time knowledge via X platform integration.
Input
$3.00
Output
$15.00
Cached input
$0.75
Source ↗
per 1M tokens
Open model
xAI
Active
Grok 4 Fast (Reasoning)
131.1K ctx Reasoning
Fast reasoning variant of Grok 4. Optimized for speed while maintaining reasoning capabilities.
Input
$0.20
Output
$0.50
Source ↗
per 1M tokens
Open model
xAI
Active
Grok 4.1 Fast (Reasoning)
131.1K ctx Reasoning
Updated fast reasoning model with improved performance and cached input support.
Input
$0.20
Output
$0.50
Cached input
$0.05
Source ↗
per 1M tokens
Open model
xAI
Active
Grok Code Fast 1
131.1K ctx
Specialized coding model from xAI. Optimized for code generation, review, and debugging.
Input
$0.20
Output
$1.50
Source ↗
per 1M tokens