Developer Tool • VS Code Extension

Eymia Local AI Workbench for VS Code

A private, local-first AI coding assistant for developers and small teams who want project-aware help, safe patch review, and local model workflows without surprise token bills.

What it does

Private AI coding help inside VS Code.

  • Local-first assistance with Ollama support and flexible OpenAI-compatible backend options
  • Project-aware Q&A with workspace indexing, file search, and codebase context
  • Explain, refactor, fix, test generation, guided workflows, and autonomous task modes
  • Patch preview, diff review, apply flow, rollback history, and safer change approval
  • MCP tool integration, connection checks, and approval-aware task execution
  • Local privacy mode, startup diagnostics, and workspace-aware controls for trust-sensitive environments
Who it is for

Built for control-first developers and teams.

  • Solo developers, freelancers, students, and indie builders using local models
  • Small software teams that want shared workflows without default cloud code exposure
  • Businesses with sensitive codebases that need approvals, auditability, and private deployment paths
Core promise: no forced cloud dependency, no surprise token bills from Eymia, and no code leaving your machine unless you choose.
How it works

A simple launch flow built around clear human approval.

  1. Connect the extensionInstall the VS Code extension, connect Ollama or a compatible backend, and choose your preferred model.
  2. Index the workspaceLet Eymia understand the project so it can answer questions, search relevant files, and plan changes with context.
  3. Review and approve editsInspect proposed diffs, approve changes, optionally run tests, and keep rollback history available when you need it.
Trust and safety

Designed to feel safer than blind autocomplete.

  • Patch review before file changes
  • Approval gates for risky actions and test execution
  • Rollback and history support for recovering from bad edits
  • Local privacy mode for users who want fewer built-in network behaviors
Pricing

Simple pricing for local-first adoption.

The pricing direction favors predictable subscriptions over token credits so users can bring their own compute and local models with less uncertainty.

$0

Free Community

For developers testing local AI coding workflows with one workspace, basic code Q&A, manual patch preview, and local-only operation.

$12

Indie Pro

For solo developers and freelancers who need commercial use, multiple workspaces, indexing, rollback history, local MCP tools, and test integration.

$25

Team

For small teams that want shared rules, reusable agents, team MCP registry plans, admin controls, audit history, and private knowledge workflows.

Private deployment is planned as a separate contact-sales path starting at $499/month for organizations that need self-hosted or local-network installation support.
Coming soon

The core product is built, and launch polish is next.

The current go-live work is focused on final patch-review QA, installed-VSIX validation, and a clean public launch story for the first release channel.

Team governance, stronger on-prem support, and broader deployment controls are part of the path after the first public launch.