Pricing
The tool is free. The full tool. Pay only when you want managed billing or a team of agents.
“We get paid based on how valuable this is to you. Not before.”
Engram
Free, forever
- Open-source CLI tool
- Skills, hooks, memory, MCP server
- Bring your own API keys
- Works with any model — Claude, GPT, Gemini, Ollama, and more
- Full power, zero limitations
- Community support
The full tool. Not a trial. Not a teaser. The real thing.
Engram Cloud
Pay as you go
- Everything in Free, plus:
- Built-in billing — one bill for 200+ models via OpenRouter
- Cloud memory sync across devices
- Transparent pricing: compute cost + ~5% margin
- No subscription — pay only for what you use
- Usage dashboard & cost tracking
Same tool, managed billing. We add ~5% to cover infrastructure. That's it.
The Lab
Coming Soon
- Everything in Cloud, plus:
- Managed AI agent team
- Pre-built skill library
- Collaboration & team features
- Priority support
For teams that want agents working alongside them. Not just a tool — a team.
The free tier isn’t limited. It’s the same tool with the same features. Cloud just adds managed billing and sync. You bring your own API keys either way — Cloud simply gives you one bill instead of managing multiple providers yourself.
Skill Marketplace
Coming SoonCreate, share, and sell skills. A marketplace for AI capabilities built by the community, for the community. Available across all tiers.
Frequently Asked Questions
There's no catch. Engram is a fully open-source CLI tool. You get skills, hooks, memory, MCP server — everything. You bring your own API keys from whatever model providers you prefer. We don't limit features, usage, or capabilities.
They're the same tool. Cloud adds two things: unified billing (one bill for 200+ models through OpenRouter instead of managing separate API keys) and cloud memory sync across your devices. That's it. No features are withheld from the free tier.
You pay the actual compute cost of the models you use, plus roughly 5% to cover our infrastructure. No subscriptions, no minimums, no surprises. If you don't use it, you don't pay. Your usage dashboard shows exactly what you're spending and where.
Any model you want. Claude, GPT, Gemini, Ollama, Llama, Mistral — Engram is model-agnostic. On the free tier, you use your own API keys. On Cloud, OpenRouter gives you access to 200+ models through a single account.
Yes. Your skills, memory, and configuration are model-agnostic. Switch from Claude to GPT to Gemini to a local Ollama model without losing anything. That's the whole point.
The Lab is coming soon. It adds a managed AI agent team, a pre-built skill library, collaboration features for teams, and priority support. Join the waitlist if you're interested.
Those are model subscriptions tied to one provider. Engram is infrastructure you own. Your ChatGPT Plus subscription doesn't give you portable skills, persistent local memory, or the ability to switch to Claude tomorrow without losing everything. Engram does.
That's what the free tier is. Engram is open source. Run it locally, use your own keys, own your data. If you ever outgrow Cloud, you can always go back to self-hosting. No lock-in, ever.
Compare Plans
| Feature | Engram (Free) | Engram Cloud | The Lab |
|---|---|---|---|
| CLI Tool | |||
| Skills, Hooks, Memory | |||
| MCP Server | |||
| Any Model (Claude, GPT, Gemini, Ollama) | |||
| Community Support | |||
| Unified Billing (200+ models) | — | ||
| Cloud Memory Sync | — | ||
| Usage Dashboard | — | ||
| Managed Agent Team | — | — | |
| Pre-built Skill Library | — | — | |
| Team Collaboration | — | — | |
| Priority Support | — | — |
Ready to own your AI infrastructure?