Cloudflare R2 - Zero Egress Object Storage for Developers

Posted on: 4/26/2026 4:14:32 AM

1. What is Cloudflare R2 and Why Should Developers Care?

Cloudflare R2 is an object storage service built on a simple but powerful philosophy: zero egress fees — you pay nothing when downloading data. This might sound minor, but it is the single biggest differentiator compared to AWS S3, Azure Blob Storage, or Google Cloud Storage, where egress fees can consume 50-80% of your total storage bill.

Imagine running a CDN serving 10TB of data per month. With AWS S3, egress alone costs $920/month. With R2? $0. That is not a small saving — it fundamentally changes how you design system architecture.

2. R2 Technical Architecture

R2 is not simply a "cheaper S3 clone." Its architecture leverages Cloudflare's massive edge network — over 330 data centers worldwide — to deliver a fundamentally different experience.

graph TD
    A["Client Request"] --> B["Cloudflare Edge Network
330+ PoPs"]
    B --> C["R2 Storage Layer"]
    B --> D["Workers Runtime"]
    D -->|"Binding"| C
    C --> E["Standard Storage
$0.015/GB"]
    C --> F["Infrequent Access
$0.01/GB"]
    D --> G["Event Notifications"]
    G --> H["Queues Consumer"]
    style A fill:#e94560,stroke:#fff,color:#fff
    style B fill:#0f3460,stroke:#fff,color:#fff
    style C fill:#4CAF50,stroke:#fff,color:#fff
    style D fill:#ff9800,stroke:#fff,color:#fff

Figure 1: Cloudflare R2 Architecture — deep integration with Edge Network and Workers Runtime

2.1. S3 API Compatibility

R2 is S3 API compatible, meaning most of your existing tools and SDKs work immediately without code changes. AWS CLI, Boto3, MinIO Client, rclone — all work with R2 by simply changing the endpoint.

# Configure AWS CLI for R2
aws configure set default.s3.endpoint_url https://<ACCOUNT_ID>.r2.cloudflarestorage.com

# Upload file — identical S3 syntax
aws s3 cp ./backup.tar.gz s3://my-bucket/backups/

# List objects
aws s3 ls s3://my-bucket/

2.2. Two Storage Tiers

R2 offers two storage classes suited for different access patterns:

3. Cost Comparison: R2 vs S3 vs Azure Blob vs GCS

This is the part that surprises most teams. Here are the actual costs for a common workload: 1TB storage + 5TB egress/month.

graph LR
    subgraph "Monthly Cost - 1TB + 5TB egress"
    A["R2: ~$20"] --- B["S3: ~$488"]
    B --- C["Azure: ~$458"]
    C --- D["GCS: ~$625"]
    end
    style A fill:#4CAF50,stroke:#fff,color:#fff
    style B fill:#ff9800,stroke:#fff,color:#fff
    style C fill:#ff9800,stroke:#fff,color:#fff
    style D fill:#e94560,stroke:#fff,color:#fff

Figure 2: Cost comparison — R2 is 20-30x cheaper than Big 3 thanks to zero egress

4. R2 + Workers: The Power of Edge Computing

R2's biggest differentiator compared to other object storage services: native integration with Cloudflare Workers. You can process, transform, and serve objects directly at the edge — no separate server, no Lambda@Edge required.

4.1. Wrangler Configuration

# wrangler.toml
name = "r2-image-service"
main = "src/index.ts"
compatibility_date = "2026-04-01"

[[r2_buckets]]
binding = "MEDIA_BUCKET"
bucket_name = "media-production"
preview_bucket_name = "media-preview"

4.2. Worker for Upload and Serve

interface Env {
  MEDIA_BUCKET: R2Bucket;
}

export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const url = new URL(request.url);
    const key = url.pathname.slice(1);

    switch (request.method) {
      case "PUT": {
        const object = await env.MEDIA_BUCKET.put(key, request.body, {
          httpMetadata: {
            contentType: request.headers.get("content-type") || "application/octet-stream",
          },
          customMetadata: {
            uploadedBy: request.headers.get("x-user-id") || "anonymous",
            uploadedAt: new Date().toISOString(),
          },
        });
        return new Response(JSON.stringify({
          key: object.key,
          size: object.size,
          etag: object.etag,
        }), { status: 201 });
      }

      case "GET": {
        const object = await env.MEDIA_BUCKET.get(key);
        if (!object) return new Response("Not Found", { status: 404 });

        const headers = new Headers();
        object.writeHttpMetadata(headers);
        headers.set("etag", object.httpEtag);
        headers.set("cache-control", "public, max-age=31536000, immutable");

        return new Response(object.body, { headers });
      }

      case "DELETE": {
        await env.MEDIA_BUCKET.delete(key);
        return new Response(null, { status: 204 });
      }

      default:
        return new Response("Method Not Allowed", { status: 405 });
    }
  },
};

5. Event Notifications — Building Automated Data Pipelines

R2 Event Notifications allow you to automatically trigger Workers when bucket data changes — similar to S3 Event Notifications but natively integrated with Cloudflare Queues.

sequenceDiagram
    participant U as User
    participant W as Upload Worker
    participant R2 as R2 Bucket
    participant Q as Cloudflare Queue
    participant P as Processing Worker
    U->>W: Upload image
    W->>R2: PUT object
    R2->>Q: Event notification
    Q->>P: Consume message
    P->>R2: GET original
    Note over P: Resize, optimize, generate thumbnails
    P->>R2: PUT processed versions

Figure 3: Event-driven pipeline — automatic image processing on R2 upload

5.1. Event Notifications Configuration

# wrangler.toml
[[r2_buckets]]
binding = "MEDIA_BUCKET"
bucket_name = "media-production"

[[queues.consumers]]
queue = "r2-events"
max_batch_size = 10
max_batch_timeout = 30

interface Env {
  MEDIA_BUCKET: R2Bucket;
}

export default {
  async queue(batch: MessageBatch<R2EventNotification>, env: Env) {
    for (const message of batch.messages) {
      const event = message.body;

      if (event.action === "PutObject" && event.object.key.startsWith("uploads/")) {
        const original = await env.MEDIA_BUCKET.get(event.object.key);
        if (!original) continue;

        const thumbnailKey = event.object.key.replace("uploads/", "thumbnails/");
        console.log(`Processed: ${event.object.key} -> ${thumbnailKey}`);
      }

      message.ack();
    }
  },
};

6. Migrating from S3 to R2

Cloudflare provides two free migration tools:

6.1. Super Slurper — Bulk Migration

Super Slurper copies all objects from S3 (or GCS) to R2. Suitable for one-time migrations or scheduled syncs.

graph LR
    A["AWS S3 Bucket"] -->|"Super Slurper"| B["Cloudflare R2"]
    C["Google Cloud Storage"] -->|"Super Slurper"| B
    D["S3-Compatible"] -->|"Super Slurper"| B
    style B fill:#4CAF50,stroke:#fff,color:#fff
    style A fill:#ff9800,stroke:#fff,color:#fff

Figure 4: Super Slurper supports migration from multiple sources

6.2. Sippy — Incremental Migration

Sippy is the smarter solution: it acts as a proxy layer. When a client requests an object:

1. R2 checks if the object exists in the bucket
2. If yes — serve from R2
3. If no — fetch from S3 source, store in R2, then serve

Result: you migrate gradually with zero downtime, objects are copied on-demand when there is actual traffic. After some time, most hot data already resides in R2.

7. Real-World Use Cases for R2

7.1. CDN Origin for Static Assets

Use R2 as origin server for website assets (images, CSS, JS). Combined with Cloudflare CDN cache, you get a complete zero-cost egress solution.

7.2. Backup and Log Storage

Database backups, application logs, audit trails — use the Infrequent Access tier at just $0.01/GB. 56% cheaper than S3 Standard.

7.3. AI/ML Dataset Storage

Training datasets are often very large and need to be downloaded multiple times. Zero egress makes R2 the ideal choice for AI teams with limited budgets.

7.4. Multi-region Content Delivery

R2 + Workers enables serving content with custom logic at the edge: A/B testing, personalization, geo-routing — all without traditional origin servers.

7.5. Presigned URLs for Secure Uploads

async function generatePresignedUrl(env: Env, key: string): Promise<string> {
  const s3Client = new S3Client({
    region: "auto",
    endpoint: `https://${env.ACCOUNT_ID}.r2.cloudflarestorage.com`,
    credentials: {
      accessKeyId: env.R2_ACCESS_KEY,
      secretAccessKey: env.R2_SECRET_KEY,
    },
  });

  const command = new PutObjectCommand({
    Bucket: env.BUCKET_NAME,
    Key: key,
    ContentType: "image/jpeg",
  });

  return await getSignedUrl(s3Client, command, { expiresIn: 3600 });
}

8. Public Buckets and Custom Domains

R2 supports public bucket access — turning your bucket into a static file server accessible directly via URL. Combined with a custom domain, you have a complete CDN origin.

# Configure public access via Wrangler
npx wrangler r2 bucket update my-bucket --public

# Access objects via public URL
# https://pub-{hash}.r2.dev/{object-key}

# Or map a custom domain via Cloudflare Dashboard
# https://assets.yourdomain.com/{object-key}

9. CORS, Multipart Upload and Advanced Features

9.1. CORS Configuration

[
  {
    "AllowedOrigins": ["https://yourdomain.com"],
    "AllowedMethods": ["GET", "PUT", "POST", "DELETE"],
    "AllowedHeaders": ["*"],
    "MaxAgeSeconds": 3600
  }
]

9.2. Multipart Upload for Large Files

R2 supports multipart upload for large files — splitting files into multiple parts and uploading in parallel. Compatible with S3 multipart API, supporting files up to 5TB.

9.3. R2 Data Catalog (Beta)

A new feature allowing you to query object metadata without listing the entire bucket. Useful for analytics and data management on large buckets.

10. Best Practices When Using R2

11. When Should You Choose R2?

12. Conclusion

Cloudflare R2 is not an "S3 killer" in every scenario — but it is the best choice for the majority of common workloads. Zero egress does not just save money; it frees you from "vendor lock-in via billing" — the phenomenon where egress costs are so high that migrating data out of a cloud becomes financially unfeasible.

Combined with Workers, Event Notifications, and Queues, R2 becomes a complete object storage + edge compute platform. Especially with the free tier of 10GB and zero egress, this is an unmissable choice for startups, side projects, and anyone looking to optimize cloud costs.

Key takeaways:

• Zero egress — 20-30x cheaper than Big 3 for read-heavy workloads
• S3 API compatible — easy migration, no SDK/tool changes needed
• Workers + R2 — native edge compute + storage, no extra services
• Event Notifications — fully serverless event-driven pipelines
• Sippy migration — move from S3 with zero downtime, zero cost

References:
Cloudflare R2 Documentation | R2 Pricing | R2 Event Notifications & Infrequent Access Announcement | R2 Event Notifications Docs | Workers R2 API Reference