AI token counter & cost calculator
Paste a prompt to count its tokens and compare the cost on GPT, Claude and Gemini models — per call and per thousand calls. See exactly how your text is split, including Lao.
Our tools run in your browser — nothing you paste or open is uploaded to us or stored by us.
- Tokens (o200k)
- …
- Characters
- …
- Words
- …
- Tokens per word
- —
Loading the tokenizer (about 2 MB, from this site)…
The token count uses OpenAI's open o200k_base tokenizer, running in your browser. Each vendor counts a little differently, so model figures are estimates: for Claude 4.7 and later we show a range, because Anthropic says its newer tokenizer produces about 30% more tokens. For an exact number before a big job, use the vendor's own token-count endpoint.
Standard prices per million tokens copied from the official pages on 2026-10-08; batch, cache and regional prices differ. Prices change — check the source: OpenAI API pricing · Claude API pricing · Gemini API pricing
| Model | $ per 1M in / out | Prompt tokens | Cost per call | Cost per 1,000 calls |
|---|---|---|---|---|
| GPT-6 Astra | $10 / $50 | 0 | $0.025 | $25.00 |
| GPT-6.1 Sol | $2 / $10 | 0 | $0.0050 | $5.00 |
| GPT-6 Luna | $0.1 / $0.5 | 0 | $0.00025 | $0.25 |
| GPT-5.6 Terra | $2 / $12 | 0 | $0.0060 | $6.00 |
| Claude Fable 5.11M token context | $10 / $50 | 0 | $0.025 | $25.00 |
| Claude Opus 5.51M token context | $4 / $20 | 0 | $0.01 | $10.00 |
| Claude Sonnet 5.51M token context | $2 / $10 | 0 | $0.0050 | $5.00 |
| Claude Haiku 5.5Higher price above 100,000 prompt tokens | $0.1 / $0.5 | 0 | $0.00025 | $0.25 |
| Gemini 3.8 FlashPrice changes on 2027-01-01 to $1.5 / $7.5 | $0.75 / $3.75 | 0 | $0.0019 | $1.88 |
| Gemini 3.1 Pro (preview)Higher price above 200,000 prompt tokens | $2 / $12 | 0 | $0.0060 | $6.00 |
| Gemini 2.5 Flash | $0.3 / $2.5 | 0 | $0.0013 | $1.25 |