Rate Limits

Request limits and quotas for the Adaline API.

Adaline enforces rate limits to ensure fair usage and platform stability. Limits apply per workspace.

Default Limits

Resource	Limit	Window
Log Trace APIs	600	per minute
Log Span APIs	600	per minute
Deployment APIs	300	per minute
All other APIs	100	per minute

These are default limits. If you need higher limits for your use case, contact [e-mail].

Handling Rate Limits

When you exceed the rate limit, the API returns a 429 Too Many Requests response:

{
  "error": {
    "code": "RATE_LIMIT_EXCEEDED",
    "message": "Rate limit exceeded. Please try again later."
  }
}

Payload Limits

Resource	Max Size
Request body	32 MB
Response body	32 MB
Span content	1 MB
Span content attachment (image)	10 MB
Span content attachment (pdf)	10 MB

Requests exceeding payload limits receive a 413 Payload Too Large response.

SDK Retry Behavior

TypeScript SDK (@adaline/client)

• The Monitor class buffers logs and flushes in batches
• Failed flushes are automatically retried with exponential backoff
• Configure flushInterval and maxBufferSize to control batching

const monitor = adaline.initMonitor({
  projectId: "your-project-id",
  flushInterval: 5,    // seconds between flushes
  maxBufferSize: 100,  // max items before forced flush
});

Gateway SDK (@adaline/gateway)

• Built-in configurable queue with automatic retry
• Exponential backoff on transient failures (429, 5xx)
• Pluggable queue backend for custom retry logic

const gateway = new Gateway({
  // Custom queue options are configured per-request or via plugins
});

Best Practices

Batch your log submissions

Use the SDK’s built-in batching (Monitor class) rather than sending individual API calls for each trace or span. This significantly reduces the number of requests.

Cache deployments locally

Use initLatestDeployment() with a refreshInterval to cache prompt deployments locally and reduce deployment fetch requests.

Implement exponential backoff

If making direct API calls, implement exponential backoff when you receive 429 responses. Start with a 1-second delay and double it on each retry, up to a maximum of 60 seconds.

Monitor your usage

Review plan limits in your workspace settings and contact support if your application needs higher limits.