LLM API Cost Calculator
Estimate the cost of one request or a month of usage. Choose a model, enter your token counts, and compare rates across providers.
Your estimate
Updates instantlyGPT-6 Luna
Based on your usage and provider-checked pricing.
Estimated monthly cost
$13.50
1,000 requests per day · 30 days
- Per request
- $0.00045
- Per 1,000 requests
- $0.4500
Reference rates
USD per 1 million tokens
- Input
- $0.10
- Cached input
- $0.01
- Output
- $0.50
Compare models
Prices by provider
USD per 1 million tokens
Explore provider rates and compare the cost of your current request across models.
OpenAI4 models
GPT-6 Luna
Selected model
Per request
$0.00045
- Input
- $0.10
- Cached
- $0.01
- Output
- $0.50
Over 272,000 input tokens: $0.20 in, $0.75 out.
GPT-6.1 Terra
Per request
$0.00450
- Input
- $1.00
- Cached
- $0.05
- Output
- $5.00
Over 272,000 input tokens: $2.00 in, $7.50 out.
GPT-6 Sol
Per request
$0.00900
- Input
- $2.00
- Cached
- $0.20
- Output
- $10.00
Over 272,000 input tokens: $4.00 in, $15.00 out.
GPT-6.1 Sol
Per request
$0.00900
- Input
- $2.00
- Cached
- $0.10
- Output
- $10.00
Over 272,000 input tokens: $4.00 in, $15.00 out.
USD per 1 million tokens. OpenAI pricing page
Anthropic4 models
Claude Haiku 5.5
Per request
$0.00045
- Input
- $0.10
- Cached
- $0.01
- Output
- $0.50
Over 100,000 input tokens: $0.50 in, $2.50 out.
Claude Sonnet 5.5
Per request
$0.00900
- Input
- $2.00
- Cached
- $0.10
- Output
- $10.00
Claude Opus 5.5
Per request
$0.0180
- Input
- $4.00
- Cached
- $0.20
- Output
- $20.00
Claude Fable 5.1
Per request
$0.0450
- Input
- $10.00
- Cached
- $0.25
- Output
- $50.00
USD per 1 million tokens. Anthropic pricing page
Google4 models
Gemini 3.5 Flash Lite
Per request
$0.00185
- Input
- $0.30
- Cached
- $0.03
- Output
- $2.50
Gemini 3.6 Flash
Per request
$0.00337
- Input
- $0.75
- Cached
- $0.075
- Output
- $3.75
From Jan 1, 2027: $1.50 in, $7.50 out.
Gemini 3.8 Flash
Per request
$0.00337
- Input
- $0.75
- Cached
- $0.075
- Output
- $3.75
From Jan 1, 2027: $1.50 in, $7.50 out.
Gemini 4 Argon
Per request
$0.00900
- Input
- $2.00
- Cached
- $0.10
- Output
- $10.00
From Dec 1, 2026: $4.00 in, $20.00 out.
USD per 1 million tokens. Google pricing page
xAI4 models
Grok 4.3
Per request
$0.00375
- Input
- $1.25
- Cached
- $0.20
- Output
- $2.50
Over 200,000 input tokens: $2.50 in, $5.00 out.
Grok 4.5
Per request
$0.00700
- Input
- $2.00
- Cached
- $0.30
- Output
- $6.00
Over 200,000 input tokens: $4.00 in, $12.00 out.
Grok 4.6
Per request
$0.00700
- Input
- $2.00
- Cached
- $0.50
- Output
- $6.00
Over 200,000 input tokens: $4.00 in, $12.00 out.
Grok 4.7
Per request
$0.00700
- Input
- $2.00
- Cached
- $0.50
- Output
- $6.00
Over 200,000 input tokens: $4.00 in, $12.00 out.
USD per 1 million tokens. xAI pricing page
Mistral4 models
Mistral Small 4
Per request
$0.00060
- Input
- $0.15
- Cached
- $0.015
- Output
- $0.60
Ministral 3 14B
Per request
$0.00050
- Input
- $0.20
- Cached
- $0.02
- Output
- $0.20
Mistral Large 4
Per request
$0.00245
- Input
- $0.70
- Cached
- $0.07
- Output
- $2.10
Mistral Medium 3.5
Per request
$0.00675
- Input
- $1.50
- Cached
- $0.15
- Output
- $7.50
USD per 1 million tokens. Mistral pricing page
DeepSeek2 models
DeepSeek V4.1 Flash 1017
Per request
$0.00120
- Input
- $0.30
- Cached
- $0.006
- Output
- $1.20
Peak rate. Off-peak: $0.15 in, $0.003 cached, $0.60 out.
DeepSeek V4 Pro
Per request
$0.00462
- Input
- $1.32
- Cached
- $0.044
- Output
- $3.96
Peak rate. Off-peak: $0.66 in, $0.022 cached, $1.98 out.
USD per 1 million tokens. Peak hours: Mon-Fri 01:00-04:00 and 06:00-10:00 UTC, excluding Chinese public holidays. DeepSeek pricing page
Prices are from models.dev as of Nov 5, 2026, checked against each provider's pricing page for the models listed in the tables. They are estimates: check the provider before you commit.
How the cost is worked out
Providers charge per million tokens, with one price for input and a higher one for output. The cost of a request is input tokens times the input price plus output tokens times the output price. Cached input is charged at the model's cache-read price when one is listed.
Some models charge more once a prompt passes a size limit, and the limit differs by provider. The calculator applies the higher rates when your input tokens are over that model's own limit.
What is not included
Batch, flex, and fast-mode pricing, regional uplifts, cache writes, hourly cache storage, and image or audio tokens are left out. A model with no direct price in the source shows as unavailable instead of using a price from somewhere else.
Where the prices come from
Prices come from the open models.dev database, kept to each provider's own entry, and are committed to this site after a review. A model can appear days after launch, and a provider can change a price before this page does. For a full price comparison with sources, read the cheapest AI API pricing compared, or browse more notes on building with AI.