Check out our new GPU Glossary! Read now
September 6, 20242 minute read
Product Updates: Rollbacks, batching, sandbox tunnels & more

Welcome to another round of Modal Product Updates! Here’s what’s new this month.

↩️ App rollbacks

Deployed something questionable? You can now instantly roll back your app to a previous version from the dashboard or the CLI.

App rollbacks

Read the docs

📚 Dynamic batching

The @batched decorator accumulates requests into dynamically-sized batches. This lets you trade off latency for throughput and GPU utilization.

Dynamic batching

Read the docs

🏝️ Sandbox tunnels and exec

Sandbox objects can now forward TCP ports to the internet. This is useful if, for example, you want to connect to a web servers running inside them. You can also exec and interface with additional processes inside an existing Sandbox.

We’re interested in improving Sandboxes for agentic use cases. Please reach out to us if you have feedback or feature requests here!

Read the docs

🔭 OpenTelemetry integration

You can now export logs from Modal to your OpenTelemetry provider. If there are more integrations you would like to see, please let us know!

Read the docs

🤑 Pricing changes

CPU cores and higher end GPUs got cheaper! Read the announcement.

Pricing changes

👩‍💻 Client Updates

Run pip install --upgrade modal to get the latest updates. Here are some of the highlights:

For a complete list, check out the changelog.

🥑 Fresh posts

Gradio image generation interface

As always, if you’ve built anything else using Modal, share it with us on the community Slack!

Ship your first app in minutes.

Get Started

$30 / month free compute