AI for chemistry work.
Plan synthesis routes, design experiments, audit protocols, analyze compounds and reactions, and move into computational chemistry when the work needs simulation.
Plan synthesis routes, design experiments, and audit protocols before the bench.
- Plan synthesis routes, design experiments, and audit protocols before the bench.
- Analyze compounds, hazards, solubility, spectra, descriptors, and similarity in one flow.
- Balance reactions, compute yields, energetics and kinetics, and predict products.
- Move from chemistry questions into notebooks and Rowan workflows when needed.
- Turn papers, protocols, molecules, and assay results into next-step decisions.
- Answers name their sources and state what could not be verified.
Used by researchers at leading universities, labs, and technical teams.
Chemistry workspace
One umbrella for chemistry work, with deeper pages when you need them.
Use this page as the entry point for the chemistry product surface: synthesis planning, experiment and protocol design, tool-backed chemistry answers, notebooks, and compute. The deeper pages explain each area without making this page carry every detail.
Synthesis and route planning
Start from a target, paper, patent, or sketch of a reaction. Vicena helps turn it into a workable route and next experimental questions.
Protocol and experiment design
Draft source-backed procedures, expose assumptions, and audit the plan against chemistry, physics, hazards, and operational constraints.
Compounds and reactions
Look up compounds and their hazards in PubChem, compute amounts and equilibria exactly, and predict reaction products with their reliability stated.
Calculations and notebooks
Every plan includes a computer with Jupyter notebooks for chemistry calculations, plotting, data analysis, PySCF, and RDKit.
Computational chemistry
When lightweight analysis is not enough, move into molecular modeling in notebooks and Rowan workflows for advanced compute.
Tool-backed answers, not unsupported guesses.
When Vicena needs a molecular weight, a boiling point, a GHS code, a reaction product, or a modeling result, it calls the right tool and keeps the evidence visible. When a tool cannot verify something, the answer says so.
Sources on every number
Molar masses and balanced equations come from exact calculations, hazards and properties from PubChem, boiling points from the Thermo library, and descriptors from RDKit. The answer names where each value came from, so you can verify it or rerun it yourself.
Gaps, not false safety
When PubChem or a dataset has no record for your compound, the result says so instead of returning an empty answer that could read as "no hazards". The agent labels anything it adds from its own knowledge as reasoning, not as a verified result.
Multi-tool answers in one prompt
A protocol audit might combine yield, balance, boiling-point, and hazard lookups. The agent chooses which to run and in what order, sees every intermediate result, and writes the answer. You describe the problem once.
Product surface
What the chemistry workspace covers.
The agent can start with a practical chemistry goal, then route through local tools, notebooks, protocol workflows, or external compute depending on what the task requires.
Protocol Drafting + Audit
Draft first, then check.
A draft separates source-backed steps from assumptions. An audit formalizes the procedure, writes the conditions each step must meet, and records findings, gaps, assumptions, and sources in a report.
Lab Calculations
Exact, unit-explicit chemistry.
Molar mass, balancing, limiting reactant and yields, solutions and dilution, pH and buffers, Ksp, ideal gas, rate laws, Arrhenius parameters, and reaction heat. Ambiguous input is refused, not guessed.
Compounds + Reactions
PubChem records and RDKit scripts.
PubChem identity, properties, and hazards; RDKit descriptors, similarity, and spectra hints; measured solubility; and reaction prediction with ReactionT5v2 and textbook rules.
Computational Chemistry
Notebook-first modeling, powered by Rowan where needed.
The AI creates a visible notebook for descriptors, pKa, conformers, DFT, docking, MD, and other molecular modeling tasks. Selected advanced workflows run on Rowan for Plus and Ultra plans, within credit limits.
What you can get done
Practical chemistry jobs, from planning to computation.
Each scenario below is a real chemistry task. Read what the agent does, what comes back, and which deeper product area it can route into.
Before the bench, know if your protocol actually works.
You pull a synthesis from a paper or design one yourself. You're about to commit time, reagents, and glassware. Vicena reads your protocol, checks the math, flags hazards and incompatibilities, and cites what it found.
What you see
A step-by-step audit showing the theoretical yield against your claim, atom balance, solvent boiling points against reflux temperatures, and the GHS hazards and published incompatibilities of each reagent. Every number names the record, calculation, or library it came from. Anything that cannot be verified is listed as a gap, not glossed over.
What it saves
Half an hour of manual SDS review, one "wait, is this yield actually possible?" embarrassment in group meeting, and the reagents you would have wasted running a protocol that could not work.
From a paper in your hand to a runnable protocol.
Papers, supporting information, and patents often bury the method in dense prose. Vicena extracts variables such as temperature, concentration, duration, and reagents, cross-references them against the source claims, and flags anything inconsistent or incomplete.
What you see
A structured protocol with every value labeled by its source, inferred values listed as assumptions, and PDF page numbers for values taken from an uploaded paper.
What it saves
The hour you would have spent reading methods sections three times because the supplementary has different numbers than the main text.
Reaction products, without guessing.
For patent-common reactions such as Suzuki, Buchwald, Heck, amide coupling, and Fischer esterification, Vicena can call a trained reaction model. For textbook reactions such as SN2, acid-base, and simple esterification, it can use deterministic rules. You get the predicted product, the method used, and a clear reliability boundary.
What you see
The predicted product as SMILES and structure, the method that produced it, and where the prediction is expected to be reliable vs. where it is not.
What it saves
Mental bandwidth on the reactions where the answer is already known, so you can spend your thinking on the ones that are not.
Drug-likeness and scaffold analysis in one prompt.
Paste a SMILES. Get LogP, MW, HBD, HBA, TPSA, rotatable bonds, Lipinski violations, and Tanimoto similarity to the reference compounds you choose.
What you see
A full descriptor table with rule-based assessment, plus similarity scores against your reference set with scaffold analysis.
What it saves
Five tabs of SwissADME plus manual fingerprint comparison, with consistent methodology and structured output.
Computational chemistry without the boilerplate.
You want descriptors, pKa, conformers, a HOMO-LUMO gap, an optimized geometry, a dipole moment, or a vibrational spectrum. Ask in plain language. Vicena creates the notebook, chooses the right local or Rowan-powered workflow, runs within budget limits, and returns the numbers.
What you see
A Jupyter notebook with the inputs, workflow UUID when Rowan is used, results with proper units, tables, plots, CSV files, and comparison against reference values where available. The notebook persists: you can modify it, extend it, or cite it in your methods section.
What it saves
The setup time for input files, polling, descriptor extraction, and plotting for a single modeling question.
Application note
Synthesis planning with Vicena.
See how Vicena turns a chemistry objective into structured synthesis planning, tool-backed checks, and a decision-ready research artifact.
Example prompts below. Click one to run it.
Example prompts
What you can ask.
Five categories, three prompts each, from quick lookups to multi-step workflows. Click any to open a fresh chat with that prompt.
Lab safety
GHS hazards, published incompatibilities, and protocol-level audit.
Green chemistry
Solvent selection with safety cross-checks. Atom economy and yield analysis.
Compound profiles
Identity, structure, physical properties, hazards, solubility, spectra, descriptors, and similarity.
Reactions
Forward products, retrosynthesis, balancing, yields, energetics, and kinetics.
Computational chemistry
PySCF notebooks on your computer, plus Rowan workflows for advanced modeling.
Every plan, Free included, has the computer and notebooks; Rowan workflows need Plus or Ultra.
Complete workflows
Describe the problem. The agent chains the tools.
Each workflow below is a real problem chemists solve every day. Click any example to run it in a fresh chat.
Audit my synthesis protocol
Catches impossible yields, unsafe pairings, and silent hazards before you go to the bench.
Plan a synthesis for a target
Candidate routes with literature backing and the reliability of each prediction.
Characterize an unknown compound
Properties, expected spectral features, and functional groups in one pass.
Evaluate drug-likeness
Lipinski assessment with similarity to known drugs and their hazards.
Find the best solvent
Solvents ranked by measured solubility, cross-checked against GHS hazards.
Run quantum chemistry
DFT, Hartree-Fock, or MP2 calculations written and executed in your notebook.
Protocol Audit
Six checks the agent runs on a protocol.
Given a protocol, the agent rewrites it as structured steps and runs the checks below. Each returns a value with its source, or says that the data is missing. The agent writes the summary from those results and lists anything it could not check.
Yield validation
Computes the limiting reactant and theoretical yield from balanced coefficients and your amounts. A claimed yield over 100% is flagged as a conservation-of-mass violation.
Atom balance
Balances the reaction from formulas and reports elements that appear on one side only, which catches atoms from nowhere in a proposed route.
Incompatibilities
Reads each reagent's reactive groups and published incompatibility statements from PubChem, with sources, so the agent can judge pairs such as bleach and ammonia or HF and glass.
Boiling points
Values from the Thermo library, compared with reflux and reaction temperatures, with vapor pressure at the stated temperature.
GHS hazards
GHS classification read from PubChem when you ask: signal word, pictograms, and hazard statements with their sources.
Missing data
When a check has no record for a compound, the audit says so explicitly. It never reports "no hazards detected" for missing data.
Who typically uses these tools.
Who it's for
Who these tools are for.
For professionals and students. For bench chemists and theorists. The barrier Vicena lowers is the same in both cases: the friction between reading about chemistry and actually doing it.
Bench chemists
Catch impossible yields and unsafe combinations before running a protocol.
Protocol audit + Compound safetyMedicinal chemists
Evaluate drug-likeness, find analogs, plan synthesis routes.
Retrosynthesis + Descriptors + SimilarityPIs & safety officers
Review student protocols at scale, with cited safety reasoning.
Protocol audit + GHS hazards + IncompatibilitiesGrad students
Predict spectra, pick solvents, understand unknown molecules.
Spectra hints + Measured solubility + Functional groupsComputational chemists
Run molecular modeling workflows in visible notebooks without wiring every SDK call by hand.
Rowan + Notebooks + Molecular workbenchAll the capabilities
The chemistry toolkit, in full.
The chemistry and molecular modeling capabilities in the catalog. Click any one for its full reference.
Chemistry
5 tools Compound Lookup & Safety
Identity, properties and hazards from PubChem
Lab Calculations
Exact stoichiometry, solutions, equilibria and kinetics
Molecule Analysis
Descriptors, similarity and spectra hints with RDKit
Reaction Prediction
Textbook transformations, products and retrosynthesis
Solubility & Thermodynamic Data
Measured solubility and tabulated properties
Open-source libraries under the hood.
Libraries used
What the tools are built on.
Open-source chemistry libraries and public databases. The specific library or database behind each answer is cited in the output.
RDKit
Industry-standard cheminformatics toolkit used at Pfizer, Novartis, and Merck.
PubChem
NIH's open chemistry database: 116 million compounds with structures, properties, and safety data.
Thermo
70,000 compounds with validated thermodynamic properties (DIPPR correlations).
PySCF
Academic quantum chemistry package for Hartree-Fock, DFT, and post-HF calculations.
Rowan
Advanced molecular modeling workflows that Vicena submits to Rowan on Plus and Ultra plans, paid from account credits.
BigSolDB 2.0
103,944 measured solubility values, each with the DOI of its source paper.
ReactionT5v2
A T5 reaction model trained on patent reactions, for product prediction and retrosynthesis.
Common questions
FAQ.
Can the AI check if my yield is physically possible?
Yes. A reaction-amounts calculation takes the balanced coefficients and your masses or moles, finds the limiting reactant, and computes the theoretical and percent yield. A claimed yield above 100% is a conservation-of-mass violation. Ask: "Is 6.2 g of product from 5 g of salicylic acid possible for aspirin synthesis?"
How does it check whether two chemicals are compatible?
It looks up each compound's PubChem safety record: GHS hazard statements, reactive groups, reactivity profile, and published incompatibility statements, each with its source. The agent weighs those records for the specific pair and conditions. A missing record is reported as missing, never as evidence that the pair is safe.
Which reaction prediction model does it use?
ReactionT5v2, a T5 model pre-trained on the Open Reaction Database and fine-tuned on USPTO patent reactions. Its published top-1 accuracy is about 97% for forward prediction and 71% for one-step retrosynthesis on those benchmarks, and lower on chemistry unlike patent reactions. Standard textbook transformations, such as esterification, amide formation, and hydride reduction, use deterministic reaction rules instead. Predictions say nothing about yield, selectivity, or safety.
Does it do retrosynthesis?
Yes, one step at a time, with the retrosynthesis model of ReactionT5v2. Give it a target and it proposes precursors; the agent combines steps and checks them against the literature. It works best for drug-like organic molecules and is less reliable for polymers, organometallics, and inorganic targets.
How is this different from SciFinder, Reaxys, or IBM RXN?
SciFinder and Reaxys are curated literature and reaction databases, and IBM RXN is a reaction predictor. Vicena combines PubChem, exact chemistry calculations, RDKit, Thermo, BigSolDB, ReactionT5v2, PySCF, and Rowan, and an AI agent chooses which to use for each question. It does not replace the full literature coverage of SciFinder or Reaxys.
Is the GHS hazard data up to date?
It is read from PubChem when you ask, with the signal word, pictograms, hazard statements, and the share of notifiers reporting each one.
What computational chemistry methods are supported?
Every plan includes a computer with notebooks for small Hartree-Fock, DFT, and MP2 calculations with PySCF, and other packages on the computer. Plus and Ultra plans can also submit Rowan workflows, such as pKa, conformers, docking, and molecular dynamics; credit and budget limits apply before each submission.
Does it calculate the Lipinski rule of five?
Yes. An RDKit script computes molecular weight, Crippen logP, hydrogen-bond donors and acceptors, TPSA, and rotatable bonds, and counts Lipinski violations.
Can it predict spectra?
It lists expected IR bands, NMR shift ranges, and MS adducts and isotope patterns from functional groups, and can suggest groups for an observed IR band. These are not quantum-mechanical predictions; use PySCF in a notebook for first-principles vibrational spectra.
What happens when there is no data for my compound?
The result says the record or dataset does not cover it, instead of returning an empty answer that could be read as safe. The agent may then use its own chemistry knowledge, labeled as reasoning rather than as a verified result.
From the co-founder
We're trying to reduce the friction between reading chemistry and doing chemistry. Professionals and students should be able to extract ideas from papers, design experiments, audit protocols, and run computational chemistry without learning every tool one by one.
The goal is to let researchers focus on the scientific question while Vicena handles the mechanics: which tool to call, what assumptions matter, what evidence is available, and where the result stops being verified.
Use Vicena for chemistry: describe the objective in natural language, let the AI pick the right tools, and keep the evidence visible from the first idea through the executed result.
This is a small step in our journey toward AI for real science, helping students and professionals do science more enjoyably, with less friction.
Try it on your next protocol.
Free to start. Paste a protocol and ask for an audit, look up a compound, or plan a reaction.