Private EU-based inference
Serve open-weight coding models from EU infrastructure with zero request retention and no training on customer data.
Zro
ПодтвердитьZro is a private inference endpoint for coding agents on EU infrastructure, with OpenAI-compatible and Anthropic-compatible requests plus launcher setup.
Overview
Zro is a private inference endpoint for coding agents. It serves open-weight coding models on EU infrastructure and is positioned for developer tools that need responsive inference without retaining request bodies.
The product combines an API layer and a launcher layer. Developers can point compatible clients at Zro directly, or use the @moonmath-ai/zro package to launch supported tools such as Claude Code and Codex CLI with temporary session configuration. The site emphasizes zero request retention, no training on customer data, and support for long-context, multi-turn coding workflows.
Features
Compatibility with common agent APIs
Use the endpoint as OpenAI-compatible chat completions or Anthropic-compatible Messages requests, which lets existing tools connect without a new API shape.
CLI launcher for coding agents
Launch supported tools through the @moonmath-ai/zro npm package with session-scoped configuration, so the launcher can set up the provider and then clean up after exit.
Long-context model support
Work with long-context coding sessions using models that advertise large context windows, including MiniMax M3 and GLM-5.2.
Launcher-managed privacy controls
Route requests through Zro's own launcher setup, which disables vendor telemetry and analytics for launcher-supported tools and only sends web-search queries when search is used.
Current and upcoming model options
Choose from a small set of live models now, with additional open coding models listed as coming soon.
Use Cases
Run an agent against a compatible endpoint
Point an existing OpenAI-compatible or Anthropic-compatible client at Zro when you want private inference without rewriting your agent stack.
Launch coding tools with session-scoped setup
Use the launcher to start Claude Code, Codex CLI, OpenCode, Hermes, OpenClaw, Pi, or Codex App with temporary session configuration and no persistent vendor telemetry.
Select a model for the workload
Choose MiniMax M3 for long-context chat-style coding sessions or GLM-5.2 for agentic coding and tool-heavy workflows, depending on the task shape.
Connect manually configured IDE agents
Set up Cursor or Cline manually when you want those tools to call Zro directly through the OpenAI-compatible base URL.
Top up inference for short-term work
Use usage packs when you need a one-time budget for experiments, prompt tuning, evals, or bursty agent runs without committing to a subscription.
Pros and Cons
Pros
- Provides private inference on EU infrastructure with zero request retention.
- Supports both OpenAI-compatible and Anthropic-compatible request shapes.
- Offers launcher-based setup for several coding agents and apps.
- Includes long-context models suited to multi-turn coding sessions.
- Has usage packs available without a subscription for flexible top-ups.
Cons
- Model choice is limited to the models currently offered on the service, with some additional options marked as coming soon.
- Some integrations rely on manual configuration rather than a one-click native connector, especially for Cursor and Cline.
- Pricing for heavier usage depends on plan spend, usage packs, and model mix rather than a fixed unlimited allowance.
FAQ
What is Zro used for?
Zro provides private inference for coding agents through an endpoint that supports OpenAI-compatible chat completions and Anthropic-compatible Messages requests. The integrations page also describes launch paths for supported tools through the Zro CLI launcher.
Can I use my existing client or agent with Zro?
Yes. The source says Zro exposes OpenAI-compatible access for chat completions and Anthropic-compatible Messages requests at /v1/messages, so existing clients can point to the Zro base URL where supported.
Which coding tools does Zro integrate with?
The integrations page documents launcher support for Claude Code, Codex CLI, OpenCode, Hermes, OpenClaw, Pi, and Codex App. Cursor and Cline are shown with manual OpenAI-compatible setup instructions.
Does Zro keep my prompts or completions?
Zro does not retain prompt or completion bodies by default after inference is processed, and customer prompts and completions are not used for training, fine-tuning, evaluations, analytics, or dataset creation.
Where does inference run?
Zro runs inference on privacy-forward EU infrastructure. The site names Finland and France as current regions.
Quick Facts
- Category
- Developer Tool
- Product type
- Private inference endpoint
- Primary use
- Coding agents and agentic coding workflows
- Platform
- Web service with CLI launcher and API
- Supported regions
- EU infrastructure; Finland and France are named
- Source domain
- zro.moonmath.ai
Альтернативы Zro
Orca
Orca is an Agent Development Environment for shipping with coding agents, running multiple CLI agents in parallel across isolated worktrees, with desktop and mobile workflows.
Firebase Studio
Firebase Studio is a web-based workspace for full-stack app development with Gemini-assisted coding, app previews, cloud emulators, collaboration, and browser deployment.
RLAMA
RLAMA is a local AI platform for building RAG systems and intelligent agents on macOS, Linux, and Windows, with HTTP API support.
AakarDev AI
AakarDev AI helps teams manage AI provider access, project setups, logs, and analytics in one dashboard with BYOK support.
AI Magicx
AI Magicx is a unified AI workspace for chat, image, video, voice, music, email and developer tasks, helping teams and creators manage multiple models in one place.
Paper
Paper is a design tool that connects canvas, code, and AI agents so teams can create, share, and ship work in one workflow. Includes desktop app and MCP access.