onllm-dev

onllm-dev / onWatch

Public

Track AI API quotas across Synthetic, Z.ai, Anthropic (Claude Code), Codex, GitHub Copilot & Antigravity in real time. Lightweight background daemon (<50MB RAM), SQLite storage, Material Design 3 dashboard. Zero telemetry.

onwatch.onllm.dev ai-api-monitoring antigravity api-monitoring claude-code codex
135
10
100% credibility
Found Feb 11, 2026 at 25 stars 5x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Go
AI Summary

onWatch is a free open-source desktop tool that monitors real-time API quota usage for AI services like Anthropic, Synthetic, and Z.ai, displaying historical trends and projections via a local web dashboard.

How It Works

1
🔍 Discover onWatch

You hear about onWatch, a simple tool that watches your AI service usage so you never hit limits unexpectedly.

2
📥 Get it running

Run one easy command to download and set it up on your computer – it works on Mac, Linux, or Windows.

3
🔗 Link your services

Tell it which AI helpers you use (like Claude or others) by adding a quick note with your access info.

4
🚀 Start watching

Launch it – it quietly runs in the background, checking your usage every minute without slowing you down.

5
📊 See your dashboard

Open your web browser to a beautiful screen showing live charts, trends, and countdowns to when limits reset.

6
🔔 Set smart alerts

Turn on emails or phone notifications to warn you before you get close to running out.

✅ Stay in control

Now you work confidently, always knowing your usage and avoiding surprise cutoffs during important tasks.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 25 to 135 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is onWatch?

onWatch is a free, open-source Go CLI tool that keeps watch on your Synthetic, Z.ai, and Anthropic (Claude Code) API quotas in real time. It runs as a lightweight background agent, polling providers every 60 seconds by default, storing historical data in SQLite, and serving a Material Design web dashboard at localhost:9211 with live trends, reset countdowns, and projections. Devs get cross-session insights like burn rates and cycle history that providers don't offer, working with tools like Cline, Cursor, or Roo Code.

Why is it gaining traction?

Unlike provider dashboards showing just snapshots, onWatch unifies multi-provider views with time-series charts, per-session tracking, and alerts via email or browser push—no telemetry, all local. The one-line curl install daemonizes on macOS/Linux, auto-detects Claude tokens, and self-updates, making it dead simple for heavy API users. PWA support and <50MB RAM hook devs tired of surprise throttling mid-coding sprint.

Who should use this?

Solo devs or freelancers burning Anthropic/Synthetic/Z.ai quotas on Claude Code, Cline, or Windsurf, needing budget projections without SaaS trackers. Small teams sharing keys for visibility into shared spend, or privacy-focused platform engineers integrating quota data into Grafana via SQLite/REST API. Budget-conscious users on $3-60/month plans tracking efficiency across sessions.

Verdict

Grab it if you're deep in these AI APIs—solid beta with comprehensive tests and docs, despite 24 stars and 1.0% credibility signaling early days. Polish the default password and add more providers for broader appeal.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.