LLM API Cost Calculator
Enter your usage and compare the monthly cost of 34 models. Prices are editable, so you can model discounts or new rates.
| Model | $ / 1M input | $ / 1M output | Monthly cost |
|---|
Standard API list prices in USD per 1M tokens, last checked November 4, 2026. Haiku 5.5 applies to prompts up to 100K tokens; GPT-6 Sol rises to $4/$15 above 272K tokens; Gemini 3.1 Pro and Grok 4.5 cost more above 200K tokens; Gemini 4 Argon doubles to $4/$20 on December 1, 2026; DeepSeek V4 is shown at peak-hour rates (off-peak is half). Kimi, GLM, Qwen, Amazon Nova and Cohere are first-party list prices and can differ on resellers such as OpenRouter or Bedrock. Amazon Nova 2 Pro is unconfirmed: sources list both $1.25 / $10 and $2.19 / $17.50, so check AWS before relying on it. Gemini 4 Argon bills thinking tokens as output. Prices change often, so confirm on the provider's pricing page. Cached-input and batch discounts (typically 50% off) are not included.