Install Inflexa.
One command. No account, no sign-up, and nothing to disclose. Inflexa is Apache 2.0, and it runs on your machine.
curl -fsSL https://inflexa.ai/install.sh | bashRecommended. Downloads a prebuilt binary.
Then start it:
inflexaYou need a container runtime.
Inflexa runs every analysis step inside an isolated Docker or Podman container. That is what lets it work without a dependency fight: R, Python, Node, and the bioinformatics stack are already inside the image. It is also a hard requirement. There is no fallback, and Inflexa will not start without one.
- Docker or Podman
- macOS, Linux, or Windows
- ~4 GB disk for the sandbox image
The first run pulls the sandbox image, which is a few gigabytes. After that it is cached, and analyses start immediately.
Check that it works.
The first command provisions the sandbox and the local database. It is the step most likely to surface a problem, so run it before you bring real data.
inflexa setupinflexa statusIf setup fails, that is a bug worth reporting rather than something to work around. Open an issue and tell us what happened.
When it does not just work.
- Docker is required but was not found
- Inflexa cannot run without Docker or Podman. Install one, make sure the daemon is running, and check that your user can talk to it (on Linux, that usually means being in the docker group).
- The first run is slow
- It is pulling the sandbox image with the R, Python, and bioconda stack inside. That happens once.
- I want to use my own model
- Inflexa works with Anthropic and with any OpenAI-compatible endpoint, including one running locally on your own hardware. Your keys stay on your machine.
- Does it phone home?
- No. There is no account and no telemetry by default. If you want traces, you can point Inflexa at your own OpenTelemetry collector. There is no Inflexa endpoint to send them to.
Now run something.
Bring a dataset you already know the answer to. It is the fastest way to find out whether you trust the tool.