Connecting...

0token

Real-time insight, proof, and sharing layer for LLM token savings. Start your agent to see real savings data.

0%
avg reduction
$0.00
saved
0
requests
0
models
pip install 0token
waiting for proxy...
What 0token does
๐Ÿ”ฅ
Token Flamegraph
See exactly where tokens go โ€” per request, per model, before vs after compression
๐Ÿ’ฐ
Live $ Meter
Real-time animated counter showing dollars saved as your agent works
๐Ÿงพ
Receipt Cards
Shareable SVG cards โ€” download, copy, post on X
๐Ÿ”Œ
MCP Server
4 tools for any MCP-compatible agent โ€” stats, recent, flamegraph, receipt
๐Ÿ“Š
Leaderboard
Cross-tool efficiency comparison โ€” $/task at held quality
โšก
Zero Config
Works with Claude Code, Cursor, Codex, any OpenAI-compatible client
How it works
Steps
1
Install: pip install 0token
2
Start: python -m zt.cli start
3
Use your agent normally โ€” 0token compresses in background
4
Watch savings on this dashboard
Quick Setup
โ–ธ Claude Code:
set ANTHROPIC_BASE_URL=http://127.0.0.1:8787
claude
โ–ธ Cursor / OpenAI:
set OPENAI_BASE_URL=http://127.0.0.1:8787/v1
โ–ธ No API key needed!
0token auto-detects your key from ~/.claude.json