Let your R&D knowledge work for you.

Starting costs nothing. A new account comes with free credits, no card is asked for, and you can open a project, upload your own files and see what happens. After that there are monthly plans, and you can also buy credits once without any subscription. Every plan, what it includes and what it costs is listed in the app, and nothing is charged until you choose one.

Credits count the work done for you, not the number of messages you send to the AI co-scientist. A short question costs little. A long analysis across many files, or a research agent working through the literature, costs more, because it is more model time and more compute. Your balance and your usage are shown in the app. Plan credits arrive again each month and do not roll over. Credits you buy on top are not tied to a month; how long they keep is set out with the plans in the app.

Nothing is deleted, and nothing is charged without you. Your projects, files and results stay as they are, and new work waits until there are credits again. You can buy credits once, or take a plan. On a plan you can also allow usage beyond the included credits, with a spending limit you set yourself, so a long run is not cut off in the middle.

You cancel in the app, and you do not have to write to anyone. Access and the included credits you hold from a plan stay until the end of the period you paid for. If you are buying as a consumer rather than for a business, you also have fourteen days to withdraw from a purchase, on the terms set out in the app. Unspent credits and the unused part of a subscription come back, credits already spent do not, and asking us to provide the credits straight away can end the right early. The withdrawal terms in the app are the ones that govern. Payments are handled through Stripe.

Both work. You can work individually in a personal workspace or collaborate with your team in a shared organization. Projects, data, results and the plan's credits are shared inside it, and roles and permissions decide who can see and change what. Plans and credits are granted per organization. The plan is one price for the organization rather than a seat-based charge per person. Biocompile gets better when you research together, so you can invite colleagues at any time, for free.

No. There is no migration project and no schema to define first. You upload the files you already have and Biocompile is ready to go, or you start from public data alone. It does not replace the system where you keep your formal records, and it does not ask your team to move off anything. It sits next to what you have and links every answer back to the file it came from. If you do want to move your data over, a founder will gladly help you do it.

It runs it, and the code is shown to you, so you can read it, correct it or reuse it. You ask the question in plain language, the analysis runs on your data, and the result is attached to the project rather than left in a chat window. If you run certain analyses regularly, you can also build reproducible pipelines in Biocompile.

No. It proposes, and you approve. Nothing is written to your project record until a person approves it, and every entry shows whether a person asserted it or the assistant proposed it. Claims are written to carry the evidence behind them and the caveats that go with it, so you can follow one back and read the passage yourself; if a claim has no source you can open, treat it as unsupported. When the answer came from an analysis, you can read the code that produced it. Changes and file activity in your organization are written to an audit trail its administrators can review.

Spreadsheets, PDFs, instrument exports, notes, protocols, papers and sequence files, in whatever state they are in. Files come in by upload today. There is no connector for Google Drive or your instruments yet, so if an automatic pull from another system is something you need, please reach out to us.

A chat assistant answers, and the answer stays in the thread. Here the question reaches your own files as well as the literature, the analysis runs on your data and shows its code, and the result attaches to the project with its sources, once you approve it. Whoever works on the project next can read it without asking you. Biocompile connects insights to unlock new knowledge you can use. Nothing is lost in a chat somewhere.

Research data in Biocompile is stored in Europe and everything uploaded is encrypted at rest. We do not train AI models on it, and we do not share it with another customer. Where a question is answered by a model provider, we rely on their business terms, under which what is sent through their interface is not used to train their models. A project opens only for the people invited to it. You can export everything, or delete your account, whenever you want, and the full detail is in the platform privacy policy.

Literature search covers PubMed, arXiv, bioRxiv, OpenAlex and Semantic Scholar, alongside patents via our privacy-friendly partner Kwintely and registered trials. More than twenty public biological databases are reachable from the same question, among them UniProt, PDB, Ensembl, NCBI, Reactome, ChEMBL, gnomAD, ClinVar and GTEx. Your own uploaded files are searched in the same pass, and every answer links back to what it came from.

Biocompile is built for life science researchers, R&D leads and biotech founders working in synthetic biology, protein design, drug discovery, immunotherapy and industrial biomanufacturing, among other domains. It fits work whose evidence is spread across spreadsheets, PDFs, instrument exports, a person's notebook and a lot of literature. Just try it on your use case and see if it works.

Biocompile is an AI-native research workspace for life science R&D, made by Biocompile, Inc. You upload the messy files you already have, ask questions about them in plain language, and the analysis runs as Python and R code you can read. Literature, data, analyses and decisions stay connected across projects, every answer carries the source it came from, and nothing enters the project record until a person approves it. It is free to start, and research data is stored in Europe.