Token & cost savings
Ways to spend fewer tokens and less money for the same answers. 15 repos, most starred first.
caveman
A rule file, a local proxy and a code middleware that cut the tokens a coding agent writes and reads. Pick it if your agent's token bill is high and you want shorter answers without changing how you work.
#2rtk
A command-line proxy that shrinks shell output before your AI coding agent reads it
#3headroom
Shrinks tool outputs, logs, and files before your AI agent sends them to the model
#4codegraph
A local code knowledge graph that lets AI coding agents look up code in one call
#5freellmapi
One OpenAI-style API that routes across dozens of free LLM provider tiers
#6context-mode
Keeps bulky tool output out of your coding agent's context window
#79router
Local proxy that routes AI coding tools to subscription, cheap, or free models
#8pxpipe
Local proxy that renders Claude Code's bulky context as images to cut input tokens
#9Switchyard
Routing library that picks which LLM handles each agent request
#10billion-context
A proxy that compresses long coding-agent chats so one session can run for weeks
#11token-optimizer-mcp
MCP server and Claude Code plugin that cuts agent token use and keeps a project graph
#12leanctx
Python wrapper that shrinks LLM prompts locally, leaving code and tool calls exact
#13magic-compact
Compress long Claude Code and OpenCode sessions without flattening them into one summary
#14modeldeck
macOS menu bar app showing how much Claude Code and Codex usage you have left
#15reckoner
Self-hosted dashboard that shows your AI and cloud account balances in one place