Reference for platform limits and quotas.
Query Limits
| Limit | Value | Notes |
|---|
| Query timeout | 30 seconds | Queries running longer are cancelled |
| Max rows returned | 10,000 rows | Results are truncated beyond this |
| Query type | SELECT only | INSERT, UPDATE, DELETE are blocked |
Connection Limits
| Limit | Value | Notes |
|---|
| Connections per pool | 5 | Per database connection |
| Idle connection timeout | 5 minutes | Idle connections are closed |
| Max connection age | 30 minutes | Connections are recycled |
| Connections per workspace | No limit | Create as many as needed |
Notebook Limits
| Limit | Value | Notes |
|---|
| Cells per notebook | 500 | Soft limit, contact support for more |
| Result size per cell | 50 MB | Large results are truncated |
| Notebooks per workspace | No limit | Create as many as needed |
Agent Limits
| Limit | Value | Notes |
|---|
| Max steps per request | 10 | Agent stops after 10 tool calls |
| Thinking budget | 8,000 tokens | Per agent turn |
| Context window | 200,000 tokens | Includes conversation history |
AI Credits
AI usage is billed in USD credits, not tokens. Each plan includes monthly credits per seat:
| Plan | Included Credits |
|---|
| Community | — |
| Professional | $10/seat/month |
| Team | $20/seat/month |
| Enterprise | $50/seat/month |
Credits reset at the start of each billing cycle. Enable on-demand usage to continue beyond included credits (default cap: $50/month).
See Billing & Usage for details on how usage is calculated.
Semantic Layer Limits
| Limit | Value | Notes |
|---|
| Models per layer | 100 | Soft limit |
| Dimensions per model | 50 | Soft limit |
| Measures per model | 50 | Soft limit |
| Relationships | 200 | Soft limit |
| Versions retained | Unlimited | All versions are kept |
File Upload Limits
| Limit | Value | Notes |
|---|
| Max file size | 100 MB | For CSV/Parquet uploads |
| Supported formats | CSV, Parquet | More coming soon |
Rate Limits
| Endpoint | Limit | Window |
|---|
| API requests | 1,000 | Per minute |
| Agent invocations | 60 | Per minute |
| Query executions | 300 | Per minute |
Rate limit headers are included in API responses:
X-RateLimit-Limit: Max requests allowedX-RateLimit-Remaining: Requests remainingX-RateLimit-Reset: When the limit resets (Unix timestamp)
Need Higher Limits?
Contact us for custom limits on Enterprise plans.