Macyou v2: AI Catalog, Deployment API, and New Pricing
Today we're launching Macyou v2 — the biggest update since we started. Three things are new: a curated AI catalog, an OpenAI-compatible API on every deployment, and a simpler five-tier pricing model. Here's what changed and why.
AI Catalog: Browse, Pick, Deploy
The new Catalog has 50+ pre-configured AI models and agent frameworks — from Llama 3.2 and Mistral to LangChain agent stacks and Stable Diffusion. Each entry lists the hardware it needs, estimated tokens/sec, and one-click deploy. No YAML, no SSH, no guesswork.
We test every catalog template on real M4 Pro hardware before publishing. Performance numbers are measured, not estimated.
OpenAI-Compatible API
Every deployment now exposes /v1/chat/completions, /v1/completions, /v1/embeddings, and /v1/models — the same endpoints your code already uses with OpenAI. Switch by changing two lines: the base URL and the API key.
Streaming, function calling, and JSON mode all work. If you use LangChain, LlamaIndex, or the OpenAI SDK, just point it at your Macyou deployment. See the API documentation for examples.
New Pricing: Five Tiers, One Clear Trade-off
We replaced the old tier names with five straightforward plans: Starter ($149/mo), Standard ($299/mo), Advanced ($599/mo), Pro ($1,199/mo), and Max ($1,999/mo). Each tier maps to a specific Apple Silicon configuration and a deployment count. Higher tier = more RAM, more compute, more concurrent deployments. That's it.
Enterprise customers with custom SLAs, dedicated clusters, or compliance needs can still reach out for a tailored plan.
What Happens to Existing Subscriptions?
If you're on a legacy plan, we're migrating you automatically on May 6. You'll get the same or better hardware at the same or lower price — no action needed. We sent an email two weeks ago with your personal tier mapping and pricing delta.
Legacy subscribers get a 12-month grandfather period: if you prefer the old pricing, contact support and we'll keep you on it through May 2027.
What's Next
We're working on GPU-accelerated templates, team workspaces, and a usage-based billing option for burst workloads. Follow the changelog for updates.