Pay for what you use
Fewer tokens, faster sessions, less time re-explaining.
Built to save you time from the start.
5k
Free operations
2 min
Typical setup time
No lock-in
MCP-native workflow
Pro
For individual developers
Founding: 50% off for life
- Full knowledge graph
- Enhanced Context with intent detection
- 100k operations/month
Elite
Graph depth: call + dataflow + typeFull graphs + Enhanced Context
Founding: 50% off for life
Team
For collaborative teams
Founding: 50% off for life
Your data stays yours
Encrypted, never used for training, and you control access. Self-hosting planned for Enterprise.
Encrypted at Rest
All data encrypted with AES-256. Your code and memories are protected.
No Training on Your Data
We never use your code or data to train AI models.
You Control Access
Workspace permissions, API keys, and team roles give you full control.
Self-Hosting (Planned)
Enterprise plans will offer self-hosting for complete data sovereignty.
Need more operations?
Buy additional operations anytime from your billing page. They never expire.
How operations work
Different operations consume different amounts of your quota based on compute intensity.
| Operation | Cost | Description |
|---|---|---|
| Memory store | 1 | Store a memory event or knowledge node |
| Memory get | 0 | Retrieve a memory event (free) |
| Memory search | 2 | Search memory with filters |
| Memory timeline | 2 | Get memory timeline |
| Memory decisions | 2 | Get decision summaries |
| Keyword search | 2 | Traditional keyword search |
| Pattern search | 2 | Regex pattern search |
| Semantic search | 5 | AI-powered semantic search |
| Hybrid search | 5 | Combined semantic + keyword search |
| Knowledge graph query | 3 | Traverse relationships and paths |
| Dependency analysis | 10 | Analyze code dependencies |
| Impact analysis | 10 | Understand change blast radius |
| Call path analysis | 10 | Trace call paths |
| Context build | 5 | Build optimized context for LLMs |
| Context enhanced | 8 | Enhanced AI context with memory |
| Context pack | 20 | Smart context pack with distillation |
| Embeddings | 3 | Generate embeddings |
| Plan | 10 | Generate development plan |
| File indexing | 1 | Index a single file |
| Session init | 3 | Initialize session with full context |
| Session remember | 1 | Store to session memory |
| Session recall | 5 | Search session memory + code |
Compare plans
| Feature | Pro | Elite | Team | Enterprise |
|---|---|---|---|---|
| Monthly operations | 25,000 | 100,000 | 200,000/seat | Custom |
| Workspaces | 3 | 10 | 50 | Unlimited |
| Projects | Unlimited | Unlimited | Unlimited | Unlimited |
| Team members | 1 | 1 | 3 minimum | Unlimited |
| Memory retention | No expiry | No expiry | No expiry | No expiry |
| All search types | ||||
| Context pack (distilled) | ||||
| Enhanced Context | Keywords + embeddings | Smart routing + semantic | Smart routing + semantic | Smart routing + semantic |
| Knowledge graphs | Limited | Full | Full (shared) | Full + org controls |
| Graph ingest tool | Lite | Full | Full | Full |
| Code graph depth | Module imports + 1-hop | Module + call + dataflow + type | Module + call + dataflow + type | Module + call + dataflow + type |
| GitHub, Slack, Notion (Jira coming soon) | ||||
| High-performance Rust MCP server | ||||
| Desktop app (Windows, Mac, Linux) | ||||
| Team sharing | ||||
| SSO | Planned | Planned | ||
| Self-hosting | Planned | |||
| Audit logs | ||||
| Support | Priority | Priority | Priority | Dedicated |
Frequently asked questions
What happens if I run out of operations?
Your service continues but API calls will return a rate limit error. You can buy more operations anytime, or upgrade your plan. We'll send you an alert when you're at 20% remaining.
Do unused operations roll over?
Monthly plan operations reset each billing cycle. However, purchased operation top-ups never expire and roll over indefinitely.
How do team seats work?
Team plan has a 3 seat minimum ($60/mo for founding members). Each seat includes 200,000 operations/month, so 3 seats get 600,000 operations pooled. Additional seats are $20/month each (founding price) with 200,000 operations each. All team members share the pooled operations and team context.
Is my code and data private?
Absolutely. Your code and data are encrypted at rest, never used to train AI models, and only accessible by you and your team. Enterprise customers can also self-host for complete data sovereignty.