What Gemini will cost us

Pick a model, say how many people use booqly and how many tokens each of them burns in a month. Prices are the standard Vertex AI rates on the global endpoint, checked on .

From 1 to 10,000.

What we send: prompts, documents, history.

What the model writes, thinking included.

Per month

Per user
Per year

The same money buys

Document scans
Chat messages

From one user to ten thousand

The same tokens per user on .

Users Per month Per year Scans a month Messages a month

Gemini models on Vertex AI

Price per 1M tokens. Requests up to 200K tokens pay the first figure, longer ones the second; ours stay far below. Pick a row to run the calculator on that model.

Model Input Cached input Output Your scenario Prices

How much one scan and one message take

Estimated on 9 October 2026 from the requests our backend builds for Gemini 3.5 Flash-Lite with medium media resolution and low thinking. A scan is one document read into fields; a message is one question in the chat with its answer. Change them when real usage tells a different story.

Document scan
Chat message