interface RetryConfig {
  maxRetries: number;
  baseDelayMs: number;
  maxDelayMs: number;
  jitterMs: number;
}

async function withExponentialBackoff<T>(
  operation: () => Promise<T>,
  config: RetryConfig = {
    maxRetries: 5,
    baseDelayMs: 1000,
    maxDelayMs: 60000,
    jitterMs: 500
  }
): Promise<T> {
  for (let attempt = 0; attempt <= config.maxRetries; attempt++) {
    try {
      return await operation();
    } catch (error: any) {
      if (attempt === config.maxRetries) throw error;

      const status = error.status;

      // Only retry on rate limits (429) or server errors (5xx)
      if (status !== 429 && (status < 500 || status >= 600)) throw error;

      // Check for Retry-After header
      const retryAfter = error.headers?.['retry-after'];
      let delay: number;

      if (retryAfter) {
        delay = parseInt(retryAfter) * 1000;
      } else {
        // Exponential delay with jitter to prevent thundering herd
        const exponentialDelay = config.baseDelayMs * Math.pow(2, attempt);
        const jitter = Math.random() * config.jitterMs;
        delay = Math.min(exponentialDelay + jitter, config.maxDelayMs);
      }

      console.log(`Attempt ${attempt + 1} failed (${status}). Retrying in ${delay.toFixed(0)}ms...`);
      await new Promise(r => setTimeout(r, delay));
    }
  }
  throw new Error('Unreachable');
}

// Usage
const response = await withExponentialBackoff(() =>
  client.chat.complete({
    model: 'mistral-small-latest',
    messages: [{ role: 'user', content: 'Hello!' }],
  })
);

class TokenRateLimiter {
  private tokensUsed = 0;
  private windowStart = Date.now();
  private readonly tokensPerMinute: number;
  private readonly windowMs = 60000; // 1 minute

  constructor(tokensPerMinute = 500000) {
    this.tokensPerMinute = tokensPerMinute;
  }

  async waitForCapacity(estimatedTokens: number): Promise<void> {
    const now = Date.now();
    const elapsed = now - this.windowStart;

    // Reset window if needed
    if (elapsed >= this.windowMs) {
      this.tokensUsed = 0;
      this.windowStart = now;
    }

    // Check if we need to wait
    if (this.tokensUsed + estimatedTokens > this.tokensPerMinute) {
      const waitTime = this.windowMs - elapsed;
      console.log(`Token limit approaching. Waiting ${waitTime}ms...`);
      await new Promise(r => setTimeout(r, waitTime));
      this.tokensUsed = 0;
      this.windowStart = Date.now();
    }
  }

  recordUsage(tokensUsed: number): void {
    this.tokensUsed += tokensUsed;
  }
}

// Usage
const rateLimiter = new TokenRateLimiter(500000);

async function rateLimitedChat(messages: Message[]): Promise<string> {
  // Estimate tokens (rough: 4 chars per token)
  const estimatedTokens = JSON.stringify(messages).length / 4;

  await rateLimiter.waitForCapacity(estimatedTokens + 500); // +500 for response

  const response = await client.chat.complete({
    model: 'mistral-small-latest',
    messages,
  });

  if (response.usage) {
    rateLimiter.recordUsage(response.usage.totalTokens || 0);
  }

  return response.choices?.[0]?.message?.content ?? '';
}

import PQueue from 'p-queue';

const requestQueue = new PQueue({
  concurrency: 5,        // Max concurrent requests
  interval: 1000,        // 1 second interval
  intervalCap: 10,       // Max 10 requests per interval
});

async function queuedRequest<T>(operation: () => Promise<T>): Promise<T> {
  return requestQueue.add(async () => {
    return withExponentialBackoff(operation);
  });
}

// Usage
const results = await Promise.all(
  prompts.map(prompt =>
    queuedRequest(() =>
      client.chat.complete({
        model: 'mistral-small-latest',
        messages: [{ role: 'user', content: prompt }],
      })
    )
  )
);

class RateLimitMonitor {
  private requestCount = 0;
  private lastReset = Date.now();
  private readonly alertThreshold: number;

  constructor(alertThreshold = 0.8) {
    this.alertThreshold = alertThreshold;
  }

  recordRequest(): void {
    const now = Date.now();
    if (now - this.lastReset >= 60000) {
      this.requestCount = 0;
      this.lastReset = now;
    }
    this.requestCount++;
  }

  checkThreshold(maxRequests: number): void {
    if (this.requestCount / maxRequests > this.alertThreshold) {
      console.warn(`Rate limit warning: ${this.requestCount}/${maxRequests} requests used`);
    }
  }

  getStats(): { requestCount: number; windowRemaining: number } {
    return {
      requestCount: this.requestCount,
      windowRemaining: 60000 - (Date.now() - this.lastReset),
    };
  }
}

import time
import asyncio
from mistralai import Mistral

async def with_retry(client, max_retries=3):
    for attempt in range(max_retries):
        try:
            return await client.chat.complete(
                model="mistral-small-latest",
                messages=[{"role": "user", "content": "Hello"}]
            )
        except Exception as e:
            if hasattr(e, 'status') and e.status == 429:
                delay = (2 ** attempt) + (random.random() * 0.5)
                print(f"Rate limited. Waiting {delay:.1f}s...")
                await asyncio.sleep(delay)
            else:
                raise
    raise Exception("Max retries exceeded")

async function processBatch<T, R>(
  items: T[],
  processor: (item: T) => Promise<R>,
  batchSize = 5,
  delayMs = 1000
): Promise<R[]> {
  const results: R[] = [];

  for (let i = 0; i < items.length; i += batchSize) {
    const batch = items.slice(i, i + batchSize);

    const batchResults = await Promise.all(
      batch.map(item => withExponentialBackoff(() => processor(item)))
    );

    results.push(...batchResults);

    // Delay between batches
    if (i + batchSize < items.length) {
      await new Promise(r => setTimeout(r, delayMs));
    }
  }

  return results;
}