Sim is the open-source AI workspace where teams build, deploy, and manage AI agents. The free Sim LLM API pricing calculator estimates cost per request, per day, and per month for 121 OpenAI, Anthropic Claude, Google Gemini, and other models, using the same rates and long-context tier rules Sim uses to cost model calls in its execution logs.
Free tool
LLM API pricing calculator
Estimate what an LLM API call costs across OpenAI, Anthropic Claude, Google Gemini, and 121 current models. Enter tokens per request, the share of input served from the prompt cache, and requests per day; Sim prices each model with the same rates and long-context tier rules it uses to cost model calls in its execution logs.
Prices last updated
What drives LLM API cost
LLM API cost is tokens times rate: input tokens at the input rate, cached input at the cached rate, and output tokens at the output rate, all quoted per one million tokens. Three things move the total.
Input and output tokens
Every LLM API call is billed per token, with separate rates for the tokens you send (system prompt, conversation history, tool results, retrieved documents) and the tokens the model generates. Output tokens usually cost several times more than input tokens, and reasoning models bill their hidden reasoning tokens as output.
Cached input
When a provider supports prompt caching, a repeated prompt prefix such as a long system prompt is read from cache at a discounted cached-input rate. GPT-6 Astra, for instance, lists $1 per 1M cached input tokens against $10 per 1M uncached. The calculator bills the cached share of input at that rate and the rest at the full input rate; models without a cached rate bill all input at the input rate.
Long-context pricing tiers
Some models charge a higher rate for the whole request once its input crosses a size threshold. For example, GPT-6 Astra bills $10 per 1M input tokens up to 272k input tokens and $20 per 1M once a request goes past that. Sim resolves the tier from all input tokens, cached and uncached, and applies it to input, cached input, and output alike.
Frequently asked questions
Multiply input tokens by the input rate and output tokens by the output rate, with both rates quoted per one million tokens, then add the two. If part of the input is served from a prompt cache, bill that part at the cached-input rate instead. Multiply the per-request cost by requests per day and by 30 for a monthly estimate. The Sim pricing calculator does this for every selected model at once.
