StartupIndex
Modal logo

Modal

Run Python in the cloud without touching infrastructure

Includes 69 simulated benchmark votes3 comments

About Modal

Modal runs Python functions on cloud CPUs and GPUs by adding a decorator. Containers are defined in Python rather than a Dockerfile, cold starts are unusually quick, and you are billed by the second, which suits batch jobs and model inference.

Comments (3)

Simulated benchmark comments are labeled below and are not endorsements from real users.

  • Isaac S.Simulated3mo ago

    The failure mode to watch: Defining the container in python instead of a dockerfile felt wrong for a day and then felt obviously correct...

  • nullrichterSimulated2mo ago

    cold starts on gpu are much faster than the alternatives i tested.

  • Sasha P.Simulated2mo ago

    Per second billing is the right model for batch work. We run nightly jobs for pennies.

Sign in to leave a comment.

  1. LlamaIndex logo

    Connect your data to language models

  2. Qdrant logo

    A vector database written in Rust

  3. Groq logo

    Inference at absurd speed

  4. Suno logo

    Make a full song from a text prompt

  5. OpenRouter logo

    One API for hundreds of language models

  6. Replicate logo

    Run open source machine learning models with an API