Self-Hosting Requirements
Run ArchiCore on your own infrastructure.
System Requirements
Minimum
| Resource | Requirement |
|---|---|
| CPU | 2 cores |
| RAM | 4 GB |
| Storage | 20 GB SSD |
| OS | Linux (Ubuntu 20.04+), macOS, Windows |
Recommended (Production)
| Resource | Requirement |
|---|---|
| CPU | 4+ cores |
| RAM | 8+ GB |
| Storage | 50+ GB SSD |
| OS | Ubuntu 22.04 LTS |
Software Requirements
Required
- Node.js 18.0 or higher
- Docker 20.10+ (for containerized deployment)
- PostgreSQL 14+ (for user data)
Optional
- Redis 6+ (for caching)
- Qdrant 1.7+ (for vector search)
- Nginx (for reverse proxy)
API Keys
ArchiCore requires external API keys for AI features:
Embeddings (Required for search)
Choose one:
- Jina AI - Get API key (recommended)
- OpenAI - Get API key
LLM (Required for AI assistant)
Choose one:
- Anthropic Claude - Get API key
- OpenAI GPT - Get API key
- DeepSeek - Get API key
Network Requirements
Outbound Connections
ArchiCore needs to reach:
| Service | Domain | Port |
|---|---|---|
| Jina AI | api.jina.ai | 443 |
| OpenAI | api.openai.com | 443 |
| Anthropic | api.anthropic.com | 443 |
| GitHub | api.github.com | 443 |
Inbound (if using webhooks)
| Port | Purpose |
|---|---|
| 80/443 | Web interface |
| 3000 | API (if not using reverse proxy) |
Database Storage
Estimate storage needs:
| Data | Size per project |
|---|---|
| Index metadata | ~1 MB per 1000 files |
| Vector embeddings | ~10 MB per 1000 files |
| Analysis cache | ~5 MB per project |
Example: A 500-file project needs ~8 MB of storage.
Scaling Considerations
Single Instance
Suitable for:
- Personal use
- Small teams (< 10 users)
- < 50 projects
Horizontal Scaling
For larger deployments:
- Load balancer in front
- Multiple ArchiCore instances
- Shared PostgreSQL
- Shared Redis
- Qdrant cluster