TessaRun

Managed generative AI infrastructure

Build the product.
Skip the plumbing.

Store media, run image and video models, manage durable generation jobs, and meter usage through one API—without provisioning GPUs or stitching the infrastructure together yourself.

Start building See how it works
One request 202 Accepted 
POST /api/v1/generations
Authorization: Bearer tr_live_...

{
  "model": "minimax-h3",
  "params": {
    "prompt": "A slow dolly through...",
    "duration": 8
  },
  "metadata": {
    "user_id": "user-42",
    "project_id": "project-7",
    "scene": "opening"
  }
}
Durable jobsFixed pricingWorkspace isolation

01 / MEDIA

Durable media storage

Upload source media and keep generated outputs, metadata, and lineage available for retrieval and reuse.

02 / MODELS

Managed image and video models

Run supported models without provisioning GPUs or managing provider-specific execution.

03 / JOBS

Durable async jobs

Submit long-running generations and receive signed lifecycle events through webhooks.

04 / USAGE

Usage and cost tracking

See each job’s fixed price and attribute usage by workspace and customer reference.

How it works

From request to durable output.

The API handles media upload, generation, job status, and output retrieval.

  1. 1
    Provide input media

    Reference a ready TessaRun asset or provide an HTTPS URL for temporary input.

  2. 2
    Start a generation job

    Choose a model, submit validated parameters, and receive a fixed price.

  3. 3
    Handle the completion webhook

    Receive a signed success or failure event at your registered HTTPS endpoint.

  4. 4
    Retrieve the output

    Download the durable asset and retrieve the parameters, price, and metadata recorded with its generation job.

For developers building generative media

Add image and video generation without building the infrastructure around it.

Start building