Tools

Token calculator

Paste text to get an estimated token count (about 4 characters per token for English) and the input cost of sending it once to each model.

Paste your prompt

0 estimated tokens · 0 characters · 0 words

DeepSeek V4 Flash$0.0000 to send once

MiMo V2.5 Pro$0.0000 to send once

DeepSeek V4 Pro$0.0000 to send once

Gemini 3.8 Flash$0.0000 to send once

Claude Haiku 4.5$0.0000 to send once

GPT-5.6 Luna$0.0000 to send once

GLM 5.3$0.0000 to send once

Qwen 3.8 Max$0.0000 to send once

Claude Sonnet 5$0.0000 to send once

Gemini 3.1 Pro$0.0000 to send once

Grok 4.7$0.0000 to send once

Kimi K3$0.0000 to send once

Claude Opus 5.5$0.0000 to send once

Claude Opus 5$0.0000 to send once

GPT-5.6 Sol$0.0000 to send once

Claude Fable 5.1$0.0000 to send once

GPT-6 Astra$0.0000 to send once

How it works

No. Each model family uses its own tokenizer; this uses the common ~4 characters per token rule for English, so treat it as an estimate.

Is the count exact?

No. Each model family uses its own tokenizer; this uses the common ~4 characters per token rule for English, so treat it as an estimate.

Does the text leave my browser?

No. The estimate runs entirely in your browser.

Try Somnus with $5 of free credit.