Skip to content

AI Protocol Generator

Generate a 12-section protocol + editable step graph from a goal.

Last updated July 2026

The AI Protocol Generator converts a natural-language research goal into a complete, structured experimental protocol. It is powered by Olto AI (our tier-selected large language models) and is available on every plan, metered against your monthly AI-call allowance.

Generation returns two synchronized artifacts: the rich prose sections (objective, hypothesis, design, materials, procedure, and so on) and an editable, drag-and-drop step graph derived deterministically from the procedure. Open the step graph in the Step Designer to reorder, branch, and refine it, or send the protocol straight to a guided test run.

Every protocol can be edited after generation. From a protocol page, click Edit in the toolbar (or Edit in Designer) to modify the title, refine any section, add or remove materials and procedure steps, change the variables, or update safety notes. Edits are versioned and recorded in your security activity log. Editing an existing protocol does not consume an AI call. Only new AI generation, refinement, or AI Improve does.

Protocol generation counts as an AI call
Each Generate request consumes one AI call from your monthly allowance. There is no separate bypass. Explorer gets 5 calls/month, Researcher 200, Lab 2,000, and Enterprise unlimited. A secondary monthly spend cap also applies. Your remaining balance is shown in Billing and Settings → Security.

Input Parameters

ParameterDescriptionImpact
Research GoalPlain-language description of what you want to achievePrimary driver of protocol content
Scientific FieldBiology, Oncology, Immunology, etc.Determines terminology and methodology conventions
Experiment TypeIn vitro cell assay, CRISPR, animal model, etc.Shapes procedural structure
EquipmentAvailable instruments in your labEnsures methods are feasible for your setup
Budget LevelMinimal ($<5k) to Institutional ($200k+)Affects reagent choices and sample sizes
Timeframe1–2 weeks to 12+ monthsDetermines experiment pacing and parallelization

Protocol Sections

Every generated protocol includes 12 structured sections:

  1. Objective: a precise statement of what the experiment will determine or measure.
  2. Hypothesis: a falsifiable prediction with direction and expected magnitude.
  3. Experimental Design: arms, n values, randomization, blinding, and primary endpoints.
  4. Materials & Reagents: specific reagents, cell lines, and consumables with catalogue numbers where applicable.
  5. Procedure: step-by-step protocol with concentrations, temperatures, and durations.
  6. Variables: independent, dependent, and controlled variables tabulated.
  7. Control Groups: positive controls, negative controls, housekeeping controls, and vehicle controls.
  8. Expected Results: predicted outcomes with statistical thresholds.
  9. Statistical Analysis: the planned statistical tests, with power analysis and sample-size justification.
  10. Quality & Reproducibility: QC metrics and acceptance criteria.
  11. Troubleshooting: common failure modes with corrective actions.
  12. Safety & Risk Notes: hazard classification, PPE requirements, waste disposal, and institutional approvals.

Quality Score

Every protocol receives four scores from 0–100:

  • Feasibility: likelihood of execution given budget, equipment, and timeframe constraints.
  • Controls: adequacy of experimental controls to support valid conclusions.
  • Reproducibility: specificity of procedural detail to enable replication.
  • Clarity: completeness and precision of instructions.

Protocol Refinements

Use the refinement buttons to iteratively improve any protocol without re-entering your parameters:

  • Reduce cost: substitutes expensive reagents and optimizes quantities.
  • Compress timeline: parallelizes steps and reduces incubation times where scientifically valid.
  • Stronger controls: adds negative, positive, isotype, and internal controls.
  • Improve reproducibility: adds QC checkpoints, tightens acceptance criteria, and specifies instrument settings.
  • Add statistical plan: includes power analysis, sample size calculations, and recommended statistical tests.
  • Add troubleshooting: documents 8–10 common failure modes with corrective actions.
  • Simplify for uni lab: adapts to equipment available in a standard academic setting.
Iterative refinement
Refinements are applied on top of the current version and saved as a new version, with a before/after snapshot. All versions are preserved and viewable from the protocol detail page. Each refinement is an AI call and counts against your monthly allowance.
Related
Was this page helpful?Send feedback