-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Welcome to TokenSense! This wiki helps you get the most out of our LLM token cost calculator.
TokenSense is a free, open-source token cost calculator for LLMs. No signup required.
Visit tokensense.app to use it right now.
git clone https://github.com/artosien/tokensense-ai.git
cd tokensense-ai
npm install
npm run devThen visit http://localhost:3000
- Paste your prompt in the input field
- Select your model (Claude, GPT-4, Llama, etc.)
- See instant cost estimates for input & output tokens
- Compare models to find the most cost-effective option
- Use for budgeting API costs before deployment
- Compare model pricing across providers
- Estimate costs for batch processing
- Check token limits before sending large prompts
- Claude 3 Opus
- Claude 3 Sonnet
- Claude 3 Haiku
- GPT-4
- GPT-4 Turbo
- GPT-3.5 Turbo
- Llama 2
- Llama 3
- Mistral 7B
- Mistral Large
More models coming soon!
No. TokenSense is local-first. Your prompts are never sent to any server.
Yes! Run the local version for completely offline usage.
Yes! The core calculator is free and open-source (Apache 2.0). Premium features coming soon.
Costs are based on official pricing from each provider. Accuracy depends on model updates - we keep it current.
Yes! You can copy results or take screenshots. Export feature coming in v1.1.
- Token counting can vary slightly by model
- We use official tokenizers where available
- Let us know in Issues
# Try:
npm install
npm run dev
# Clear cache if issues persist:
rmdir .next -r -force
npm run devCheck Issues or submit an update via PR.
Found a bug? Want to add a model?
See CONTRIBUTING.md in the main repo.
- GitHub Issues: Report bugs
- Discussions: Ask questions in GitHub Discussions
- Twitter/X: @artosien
Last updated: March 2026