• 9+ LLM Providers: OpenAI, Anthropic (Claude), Google Gemini, Groq, DeepSeek, Mistral, xAI Grok, Together AI, AWS Bedrock
• Local Model Support: Full Ollama integration for privacy-first development
• Per-Feature Configuration: Set different models for chat, autocomplete, and agent tasks
• API Key Management: Secure credential storage with VS Code’s SecretStorage

Mode	Purpose
Agent	Autonomous code writing and file manipulation
Ask	Read-only exploration with web search
Plan	Design-first with human approval gates
Debug	Structured debugging with auto-cleanup
Council	Multi-agent deliberation for complex decisions

• Background Agent Manager — Monitor and control multiple concurrent agents
• One-Shot Tasks — Quick autonomous task completion from a popup
• Configurable Presets: Careful, Balanced, or Aggressive approval levels
• Task Queuing — Auto-detects CPU cores for optimal parallelism

Agent Backends:

Backend	Purpose
TurboLight	Fast, lightweight baseline
AutoAgent	Multi-agent orchestration with web browsing and memory
DeepCode	Research paper-to-code transformations
FastAgent	Deep research with GUI automation

A unique approach to complex decisions:

• Create custom councils with specialized AI agents
• Technical review, architecture, and security review templates
• Multi-round deliberation with session history
• Get multiple AI perspectives before making a decision

Text-to-Web:

• Generate full-stack web apps from descriptions
• Framework support: React, Vue, Svelte, Next.js, Nuxt, Astro
• UI libraries: Tailwind, shadcn/ui, Material UI, Chakra, Bootstrap

Text-to-Backend:

• Production-ready API services
• Frameworks: Express, Fastify, NestJS, FastAPI, Django, Flask
• Database: PostgreSQL, MySQL, MongoDB, SQLite, Redis
• Auth: JWT, OAuth2, API Key, Session-based
• Auto-generated Docker + OpenAPI/Swagger docs

Paper-to-Code / Spec-to-Code:

• Parse research papers and generate implementations
• Feed in technical specifications, get working code
• From RFC to implementation, from design doc to codebase

• Three profiles: Conservative, Balanced, Permissive
• Approval timeouts with auto-reject
• Audit logging and history
• Sub-agent permission inheritance

Sandbox Isolation:

Option	Description
Local	No isolation (fastest)
Git Worktree	Branch-level isolation
Docker	Full container isolation
Auto	Intelligent selection per task

All AI edits are held for human review:

• Changes appear in Source Control panel
• Accept/reject individual changes or batches
• Diff viewer for detailed inspection
• Group changes by file

Already using CLI agents and coding assistants? They work with Calliope:

• Amplifier — Our official CLI agent enhancement layer
• Claude Code — Anthropic’s terminal coding assistant
• Gemini CLI — Google’s command-line AI tool
• Codex CLI — OpenAI’s terminal assistant
• Any MCP-compatible tool — Model Context Protocol support

Use your existing workflow. Calliope provides the environment, governance, and infrastructure.

We’ve built a clean, private development environment:

• All telemetry disabled — No Microsoft telemetry, no phone-home
• No nagware — Clean interface without promotional content
• Zero data collection — We don’t track your code or behavior
• Local-first — Use Ollama for completely air-gapped development
• Your API keys stay local — Nothing leaves your machine except what you explicitly send to your AI provider

Built on VSCode’s open source core, stripped of the surveillance.

• 50+ Commands for all major operations
• Context-Aware Suggestions — includes current file, imports, workspace structure
• Inline Chat — right-click context menu and keyboard shortcuts
• 100+ Configurable Settings for fine-tuned control
• Completions Mode — Real-time suggestions as you type
• Chat Mode — AI paired programming through conversation
• Your Extensions Work — Compatible with existing VS Code extensions

Platform	Status
Desktop (macOS, Linux, Windows)	Full support
Web browser	Full support
VS Code Remote & SSH	Full support
GitHub Codespaces	Full support
JupyterHub	Native integration

• JupyterHub Native — Multi-user support with OAuth
• Docker Container — All-in-one deployment (Ports 8070/8071)
• DockerSpawner Compatible — Enterprise orchestration
• Health Check Endpoints — /api/status, /api/services

Select a competitor below to see how Calliope AI IDE compares:

• Maximum flexibility — 9+ LLM providers, local models, BYOK
• Real autonomy — Background agents that work while you focus
• Multiple perspectives — Council Mode for complex decisions
• Privacy — Zero telemetry, local-first, air-gapped capable
• Enterprise-ready — JupyterHub, Docker, governance integration
• Free — No subscription, bring your own API keys

Calliope AI IDE isn’t just adding AI to an editor. It’s rethinking how developers and AI work together — with multiple modes, multiple agents, and you always in control.