100 AI Agents is a self-hosted dashboard with a built-in engine. Prompt local AI, command up to 100 agents at once on your own device, routed by two commanders, schedule and post automatically, and watch it all live — powered by Ollama models that are free and run entirely on your own device.
Eleven live panels. One local engine. Two commanders. No cloud, no subscription.
Ask your agents anything and it routes to your local model at 127.0.0.1:11434. Free models, on your device, with engine state as context.
Run up to 100 agents at once, on your own device, across 14 squads. Two commanders read each task and pick the best model for it — coder, large, or fast. Save groups; pause/resume/kick workers.
Flip it on and the engine generates and posts on its own cadence — no AI in the loop. Pause anytime.
Heartbeat, live-feed, video-post, audit-package and more — cadences run inside the engine and report their last fire and result.
Today's post counts across X, Telegram, TikTok, YouTube and Discord, plus the full posting schedule at a glance.
Tracks your render queue and lists the clips in your videos folder — wired for a perfect-quality renderer with 9:16, 16:9 and 1:1 outputs.
Green/red status for Discord, Telegram, X, YouTube, TikTok, Reddit and LinkedIn — read from local credential files that never leave your machine.
Every agent with role, squad and current task, plus a done-by-squad leaderboard that updates in real time.
One-click start and boot-autostart for all three. Pick your OS at checkout and get the matching build.
Two free installs, then double-click. Everything runs locally, on any OS.
Grab Node.js and Ollama — both free. Run the included installer to pull a free starter model to your device.
Double-click Start 100 AI Agents (Windows, macOS, or Linux). The engine boots and your browser opens to localhost:8787.
Prompt models, command agents, schedule posts. Turn on autostart so it's always ready when your computer is.
100 AI Agents runs on Ollama — open-source models you download once and keep on your own computer. There's no usage bill, no API key, and nothing is sent to the cloud. Your prompts, agents and credentials stay on your machine. Run it completely offline if you want.
A complete, ready-to-run kit for your OS — plus everything you need to sell it.
No tiers, no subscription. Pay once, pick your OS, download instantly.
Windows, macOS, and Linux. You choose your OS at checkout and on the download page, and get the matching build with the right one-click start and boot-autostart scripts. The dashboard and engine themselves are identical across platforms.
100 AI Agents runs on local AI, so you must install Ollama (free) and download multiple local models onto your own device — for example a small fast model for quick jobs, a coding model, and a larger model for the two commanders. It does not use cloud models and comes with no AI subscription. Because everything runs on your own hardware, performance depends on your CPU, RAM and GPU, and larger models need a stronger machine — so results may vary.
Yes. The two things it needs — Node.js and Ollama — are both free. Ollama's AI models are free and downloaded once to your own device, so there's no API bill and no subscription. You pay $179 once for the 100 AI Agents itself. (Note: the models run on your hardware, so a capable machine is recommended for the larger models.)
Everything is driven by a single CONFIG block in the dashboard (brand, squads, agents, colors) plus engine/config.js for behavior (models, commanders, cadences). The easiest way to change anything is to ask Claude (claude.ai or the Claude desktop app) to make the edits for you — just describe what you want and point it at the files. And if any issues come up, Claude can lead the way on troubleshooting too.
Two commanders (CMD-01 and CMD-02) sit above your agents. When you leave the model on Auto-route, a commander reads the task and picks the best installed model for it — a code model for coding, the largest model for deep work, a fast model for quick jobs. Up to 100 agents can run at once.
No. The engine runs locally and the AI runs through Ollama on your machine. Prompts, agent tasks and credentials stay on your computer. You can run it fully offline.
Yes — each build includes an autostart script for its OS (Windows Startup, macOS launchd, Linux systemd) so the engine launches when your computer does. The dashboard is then always ready at localhost:8787.
Checkout runs through Stripe; afterward buyers land on a download page that automatically delivers the single build for the OS they selected at checkout.
All sales are final — there are no refunds. This is an instantly-downloadable digital product, so every purchase is final. Please review the requirements above (you need Ollama + multiple local models, and results are hardware-dependent) before you buy.