mimOE Studio

The secure on-device AI runtime, in your pocket. Interact with local AI models, manage agents, inspect runtime metrics, and discover nodes across the agentix mesh.

mimOE Studio is the visual native interface for mimOE, the universal AI runtime that runs on devices, gateways, on-premises infrastructure, and the cloud. Studio lets you manage your runtime, interact with local models, and inspect what's happening under the hood, all from a single app.

• WHAT YOU CAN DO • Interact with and test your local AI models A multi-session chat interface connects directly to loaded models, router aliases, or deployed agent mims. Streaming responses, thinking block rendering, per-chat system prompts, and full token usage stats, all built in. • Manage mims, models, and routing Deploy and inspect mims (the micro intelligence modules that carry agent logic as serverless WASM microservices). Load your AI models in GGUF and ONNX formats. • Discover authorized nodes across your agentix mesh See every mimOE node visible on your local, proximity, and account networks, with services, roles, hardware characteristics, and available models per node. Open a chat session with a model on a nearby node directly from the discovery view. • Inspect traces without infrastructure The built-in trace viewer surfaces Gantt timelines, span trees, routing decisions with candidate scores, LLM content, and token analytics per request. No Prometheus, Grafana, or Jaeger required. • Monitor live runtime metrics See which models are loaded, their token throughput, context utilization, and routing configuration, alongside live metrics for uptime, total requests, and data written.

• WHO IT'S FOR mimOE Studio is for software engineers, AI/ML engineers, and platform engineers responsible for building, deploying, and operating mimOE across devices and environments. If you're new to mimOE, Studio is the fastest way to see the platform in operation and understand how the pieces fit together, without writing a single command.

• WORKS WITH YOUR EXISTING CODE mimOE provides an OpenAI-compatible API. Change the base URL and your existing code works. No rewriting, no new SDKs, no migration. Studio lets you validate that drop-in compatibility directly from the interface.

ABOUT mimOE mimOE is a universal inference execution engine and agentix-native operating environment. Lightweight at 10-20 MB, cross-platform, and OS-agnostic, it is purpose-built for running AI on devices, gateways, on-premises infrastructure, and public clouds. It provides built-in inference execution for both generative AI models and predictive models, alongside an API gateway, MCP gateway, node and service discovery, observability, multi-cloud broker, and zero-trust security, all available to any developer building on mimOE.

What you bring to mimOE is intelligence. What mimOE gives back is operability at scale.

Learn more at https://developer.mimik.com