Vercel Storage: Edge Config vs KV vs Blob Storage Decision Guide

Vercel offers three storage solutions: Edge Config, KV, and Blob. Pick wrong and you'll face latency issues, cost overruns, or both. Each targets different use cases with distinct performance trade-offs.

Edge Config: Global Configuration Store

Best for: Application config, feature flags, A/B test variants, routing rules.

import { get } from '@vercel/edge-config';

export default async function middleware(request: NextRequest) {
  // Read config at edge (0-5ms globally)
  const maintenanceMode = await get('maintenance');

  if (maintenanceMode) {
    return new Response('Maintenance', { status: 503 });
  }

  return NextResponse.next();
}

Performance characteristics:

• Read latency: 0-5ms globally (edge cache)
• Write latency: 30-60 seconds (global propagation)
• Size limit: 500KB total
• Cost: $20/month flat rate

Edge Config gotchas: Updates aren't instant—expect 30-60 second propagation. Don't use for real-time config changes or user-specific data. The 500KB limit includes all keys combined.

KV: Redis for Edge Functions

Best for: Session storage, caching API responses, rate limiting, real-time counters.

import { kv } from '@vercel/kv';

export async function GET(request: Request) {
  const sessionId = request.headers.get('x-session-id');

  // Check session in KV (sub-10ms from edge)
  const session = await kv.get('session:' + sessionId);

  if (!session) {
    return Response.json({ error: 'Invalid session' }, { status: 401 });
  }

  // Increment page views with atomic operation
  const views = await kv.incr('views:' + sessionId);

  return Response.json({ session, views });
}

Performance characteristics:

• Read latency: 1-10ms (depends on edge location)
• Write latency: 1-10ms (immediate consistency)
• Size limit: 100MB per key, unlimited total
• Cost: $0.30 per 100K requests + $0.50/GB storage

KV gotchas: Operations are billed per request—batch with kv.pipeline() for multiple operations. TTL (expiration) is approximate, not precise. KV is eventually consistent across regions.

Blob: File Storage and CDN

Best for: User uploads, images, PDFs, video files, static assets.

import { put } from '@vercel/blob';

export async function POST(request: Request) {
  const formData = await request.formData();
  const file = formData.get('file') as File;

  // Upload to Blob with automatic CDN distribution
  const blob = await put(file.name, file, {
    access: 'public',
    addRandomSuffix: true,
  });

  return Response.json({
    url: blob.url,           // CDN URL for serving
    downloadUrl: blob.downloadUrl, // Direct download
    size: blob.size
  });
}

Performance characteristics:

• Upload latency: 100-500ms (depends on file size)
• CDN serve latency: 10-50ms globally
• Size limit: 500MB per file
• Cost: $0.15/GB storage + $0.40/GB transfer

Blob gotchas: Files are immutable once uploaded—you can't update, only delete and re-upload. The CDN cache can take 60+ seconds to update globally. Use addRandomSuffix to prevent naming conflicts.

Decision Matrix

Choose based on data characteristics:

// Configuration data (infrequent writes, global reads)
const config = await get('feature-flags');  // → Edge Config

// Session data (frequent reads/writes, user-specific)
const session = await kv.get('user:123');    // → KV

// Files (uploads, static content)
const fileUrl = await put('image.jpg', file); // → Blob

Anti-patterns to avoid:

• Using Edge Config for user data (doesn't scale, slow writes)
• Using KV for large files (expensive, not optimized for file serving)
• Using Blob for configuration (no atomic reads, CDN cache lag)

Cost optimization tips: Edge Config becomes cost-effective with high-frequency config reads. KV costs scale with request volume—cache locally when possible. Blob transfer costs matter for large files—optimize with image compression and proper caching headers.