CogniChem runs computational chemistry and biology tools (docking, structure prediction, ADMET prediction, quantum chemistry, and more) on managed cloud hardware. There is nothing to install: you pick a tool or a workflow, send your inputs, and download the results.
Ways to use CogniChem
| Interface | Best for |
|---|---|
| The app | Running tools and workflows from the browser |
| The Assistant, in the app | Describing a problem in plain language. It plans the runs and prices them, and nothing runs until you approve. |
| The REST API | Scripts and pipelines. See the API reference. |
| The Python SDK | The whole API from Python: pip install cognichem-client |
All four share one account, one wallet, and the same tools.
Workflows
A workflow chains tools into a pipeline, for example docking a compound library and then re-predicting the best poses with Boltz-2. Start from a template in the workflow gallery and run it as is, or change any step first.
Jobs and pricing
Each tool run is a job on the hardware you choose (CPU or a GPU class). Submitting a job reserves its estimated cost from your prepaid wallet. When the job finishes you pay for the actual runtime, and the rest of the reservation is released. Every tool page lists typical runtimes and the price per run on each plan.
Next steps
- Create an account and an API key.
- Run your first job: estimate, submit, wait, and download.
- Run your first workflow from a ready-made template.
- Use the Assistant to plan runs in plain language.
The concepts explain how jobs, workflows, and billing work, and the API reference lists every endpoint.