Who on your team shipped more with less? Which repo is silently eating your AI budget? Edgee Teams gives coding assistants their missing dashboard. Invite your team, connect GitHub, and every session gets tracked, attributed, and ranked. Compare compression ratios across developers. Share session stats publicly or keep them private. Climb the monthly leaderboard and claim the title of biggest token spender. Built for teams shipping with Claude Code, Codex, and other agents.

Free

Launch tags:Productivity•Developer Tools•Artificial Intelligence

Launch Team

Pioneer — Fine-tune any LLM in minutes, with one prompt

Fine-tune any LLM in minutes, with one prompt

Promoted

Edgee

Maker

📌

Hey Product Hunt 👋 Sacha here, co founder of Edgee. For our past two launches we showed what compression does at the individual level: Claude Code Compressor extending Pro limits by 26.2%, Codex Compressor cutting costs by 35.6%. Great for solo devs. But then we started talking to teams running Claude Code, Codex, and other agents across 10, 30, sometimes 100 developers. Every team told us the same thing: "We have no idea what's happening." No attribution, no visibility, no idea which repo is burning the budget, no way to know if someone just ran a $200 session by accident. So we built Edgee Team. Invite your team, and every coding assistant session gets tracked, attributed, and ranked. Connect your GitHub org, and sessions auto-link to the repos and PRs they worked on. Each session generates a detailed dashboard: requests, tokens, compression ratio, cost, and a debug view showing exactly what Edgee compressed away. The fun part: every developer gets a shareable public profile with their stats, and there's a monthly leaderboard ranking who used the most tokens. Climb it and claim the "biggest token spender of the month" title. Or don't, and stay private. Your call, per session and per user. What this unlocks: → CTOs finally see where their AI budget goes → Teams can benchmark who ships more with fewer tokens → Devs get bragging rights on their compression game → Everything still runs through the same compression engine, so you keep the savings from the previous launches Free to try, no credit card. Connect your GitHub in 30 seconds. Would love to hear what you'd want us to track next. Per-branch attribution? Slack digest of team usage? Drop ideas in the comments.

Report

2d ago

@sachamorard Does compression hold up for non-English prompts? Thinking CJK specifically, tokenizers already split those into way more tokens per character.

Report

5h ago

Edgee

Maker

@whetlan yes, no problem, with any languages. Compression happens mainly on tool results

Report

4h ago

Previous Edgee Launches

Edgee Codex CompressorUse Codex at 35.6% lower costs

Launched on April 12th, 2026

Edgee Claude Code CompressorExtend Claude Pro's limit by 26.2%

Launched on March 22nd, 2026

EdgeeThe AI Gateway that TL;DR tokens

Launched on February 12th, 2026

Forum Threads

p/edgee

•

3mo ago

Token Compression for LLMs: How to reduce context size without losing accuracy

Hey, I'm Sacha, co-founder at @Edgee

Over the last few months, we've been working on a problem we kept seeing in production AI systems:

LLM costs don't scale linearly with usage, they scale with context.
As teams add RAG, tool calls, long chat histories, memory, and guardrails, prompts become huge and token spend quickly becomes the main bottleneck.

So we built a token compression layer designed to run before inference.

View all

Edgee

The AI Gateway that TL;DR tokens

The AI Gateway that TL;DR tokens

Edgee Team

Previous Edgee Launches

Forum Threads

Token Compression for LLMs: How to reduce context size without losing accuracy

Previous Edgee Launches

Forum Threads

Token Compression for LLMs: How to reduce context size without losing accuracy

What's great

What needs improvement

What's great

What needs improvement

vs Alternatives

What's great

What needs improvement

What's great

What needs improvement

vs Alternatives