AI Models Pricing
AI Models Pricing comparison for input, output, and total token costs across GPT, Claude, Gemini, Llama, and other LLMs. Estimate API costs by usage.
Available Models
418 models| Input Modalities | |||||||
|---|---|---|---|---|---|---|---|
| 128,000 | $0.03 | $0.15 | $0.18 | Inference Net | |||
| 128,000 | $0.05 | $0.23 | $0.28 | Inference Net | |||
| 1,050,000 | $10.00 | $50.00 | $60.00 | OpenAI | |||
| 1,050,000 | $2.00 | $10.00 | $12.00 | OpenAI | |||
| 1,050,000 | $2.00 | $12.00 | $14.00 | OpenAI | |||
| 1,050,000 | $0.20 | $1.20 | $1.40 | OpenAI | |||
| 1,000,000 | $5.00 | $30.00 | $35.00 | Sakana | |||
| 1,000,000 | $2.00 | $6.00 | $8.00 | Sakana | |||
| 131,072 | $0.06 | $0.18 | $0.24 | Inclusionai | |||
| 1,048,576 | $0.15 | $0.60 | $0.75 | DeepSeek | |||
| 260,000 | $0.04 | $0.15 | $0.19 | Inception | |||
| 1,050,000 | $10.00 | $50.00 | $60.00 | OpenAI | |||
| 1,050,000 | $5.00 | $25.00 | $30.00 | OpenAI | |||
| 1,050,000 | $10.00 | $50.00 | $60.00 | OpenAI | |||
| 1,050,000 | $5.00 | $25.00 | $30.00 | OpenAI | |||
| 1,000,000 | $2.00 | $6.00 | $8.00 | Qwen | |||
| 1,048,576 | $0.10 | $0.20 | $0.30 | Meta | |||
| 1,048,576 | $1.25 | $4.25 | $5.50 | Meta | |||
| 1,048,576 | $0.75 | $3.75 | $4.50 | ||||
| 1,048,576 | $0.375 | $1.875 | $2.25 | ||||
| 1,000,000 | $10.00 | $50.00 | $60.00 | Anthropic | |||
| 1,000,000 | $5.00 | $25.00 | $30.00 | Anthropic | |||
| 131,072 | $0.06 | $0.25 | $0.31 | Ibm Granite | |||
| 1,048,576 | $0.834 | $2.501 | $3.335 | Tencent |
FAQ
Frequently asked questions about LLM pricing
Understand token pricing, cost estimates, and how to compare AI models.
What does M mean in the token calculator?
M means one million tokens. Enter 1 for 1,000,000 input or output tokens; the cost estimates update instantly.
How is LLM API cost calculated?
The estimate multiplies input tokens by the model's input rate and output tokens by its output rate, then adds both amounts. Provider-specific fees, discounts, caching, tools, images, and other add-ons may be billed separately.
Why are input and output token prices different?
Providers often price generated output higher because generation requires sequential compute. Rates vary by model and provider, so compare both using your expected workload.
Where does the pricing data come from and how often is it updated?
Model specifications and prices are retrieved from the upstream model catalog and refresh on a six-hour cache cycle when the source service is available. Rates can change, so verify the provider's current price before purchasing.
How do I choose the most cost-effective AI model?
Set your expected input and output usage, filter by provider, and sort the cost columns. Also compare context length, supported modalities, and task quality—the lowest token rate is not always the lowest total cost.