Ideogram Rate Limits

Ideogram Rate Limits

Overview

Handle Ideogram's rate limits with exponential backoff, request queuing, and concurrency control. Ideogram enforces a default limit of 10 in-flight requests (concurrent, not per-minute). Image generation takes 5-15 seconds per call, so this limit can be hit quickly during batch operations.

Prerequisites

• IDEOGRAM_API_KEY configured
• Understanding of async patterns
• p-queue npm package (optional, for queue-based approach)

Ideogram Rate Limit Model

Instructions

Step 1: Exponential Backoff with Jitter

async function withBackoff<T>(
  operation: () => Promise<T>,
  config = { maxRetries: 5, baseMs: 1000, maxMs: 30000, jitterMs: 500 }
): Promise<T> {
  for (let attempt = 0; attempt <= config.maxRetries; attempt++) {
    try {
      return await operation();
    } catch (err: any) {
      if (attempt === config.maxRetries) throw err;

      const status = err.status ?? err.response?.status;
      // Only retry on 429 (rate limited) or 5xx (server error)
      if (status && status !== 429 && status < 500) throw err;

      const exponential = config.baseMs * Math.pow(2, attempt);
      const jitter = Math.random() * config.jitterMs;
      const delay = Math.min(exponential + jitter, config.maxMs);

      console.warn(`Rate limited (attempt ${attempt + 1}/${config.maxRetries}). Waiting ${delay.toFixed(0)}ms`);
      await new Promise(r => setTimeout(r, delay));
    }
  }
  throw new Error("Unreachable");
}

Step 2: Concurrency-Limited Queue

import PQueue from "p-queue";

// Ideogram allows 10 in-flight -- use 8 to leave headroom
const ideogramQueue = new PQueue({ concurrency: 8 });

async function queuedGenerate(prompt: string, options: any = {}) {
  return ideogramQueue.add(async () => {
    const response = await fetch("https://api.ideogram.ai/generate", {
      method: "POST",
      headers: {
        "Api-Key": process.env.IDEOGRAM_API_KEY!,
        "Content-Type": "application/json",
      },
      body: JSON.stringify({
        image_request: { prompt, model: "V_2", ...options },
      }),
    });

    if (response.status === 429) {
      throw Object.assign(new Error("Rate limited"), { status: 429 });
    }
    if (!response.ok) throw new Error(`Generate failed: ${response.status}`);
    return response.json();
  });
}

// Process 50 prompts safely -- queue manages concurrency
const prompts = Array.from({ length: 50 }, (_, i) => `Design variant ${i + 1}`);
const results = await Promise.all(prompts.map(p => queuedGenerate(p)));

Step 3: Token Bucket Rate Limiter

class TokenBucket {
  private tokens: number;
  private lastRefill: number;

  constructor(
    private maxTokens: number = 10,
    private refillRate: number = 1, // tokens per second
  ) {
    this.tokens = maxTokens;
    this.lastRefill = Date.now();
  }

  async acquire(): Promise<void> {
    this.refill();
    if (this.tokens > 0) {
      this.tokens--;
      return;
    }
    // Wait for next token
    const waitMs = (1 / this.refillRate) * 1000;
    await new Promise(r => setTimeout(r, waitMs));
    this.refill();
    this.tokens--;
  }

  private refill() {
    const now = Date.now();
    const elapsed = (now - this.lastRefill) / 1000;
    this.tokens = Math.min(this.maxTokens, this.tokens + elapsed * this.refillRate);
    this.lastRefill = now;
  }
}

const bucket = new TokenBucket(10, 1);

async function throttledGenerate(prompt: string) {
  await bucket.acquire();
  return queuedGenerate(prompt);
}

Step 4: Batch with Progress Tracking

async function batchGenerate(
  prompts: string[],
  onProgress?: (done: number, total: number) => void
) {
  const results: any[] = [];
  const errors: { prompt: string; error: Error }[] = [];

  for (let i = 0; i < prompts.length; i++) {
    try {
      const result = await withBackoff(() => queuedGenerate(prompts[i]));
      results.push(result);
    } catch (err) {
      errors.push({ prompt: prompts[i], error: err as Error });
    }
    onProgress?.(i + 1, prompts.length);
  }

  console.log(`Batch complete: ${results.length} success, ${errors.length} failed`);
  return { results, errors };
}

Error Handling

Output

• Reliable API calls with automatic retry on 429
• Concurrency-controlled request queue
• Token bucket for sustained throughput
• Batch processing with progress and error tracking

Examples

scope=sandbox-generation; requested=100; limited=3; deferred=3; retry=v2; idempotent=pass; output_retention=none; rollback=limits-r7 proves bounded handling without retaining prompts or images.

Resources

• Ideogram API Overview
• p-queue
• Enterprise limits: partnership@ideogram.ai

Next Steps

For security configuration, see ideogram-security-basics.