Skip to content
AI co-scientist for computational biology

Frontier-grade computational biology on the models you run yourself.

Inflexa plans, runs and records your analyses with any model, including open weights on your own GPUs. Its Knowledge Plane grounds every plan in published methods, so an open model plans like a frontier one. Run your own model and your data and prompts stay on your network.

Free to start: the open-source CLI, or the Cloud Personal plan.

Pick a question
localhost · Inflexa Pro

Question: Which genes and pathways distinguish responders to pembrolizumab in GSE78220?

Plan: Anti-PD-1 responder DE and GSEA. Steps: Cohort assembly, Responder DE, Immune and stromal mix, GSEA on hallmark sets, Survival association.

Findings: Gene selection repeats inside every fold; Recorded sex is checked against the transcriptome.

Sessions built from public datasets: GSE78220, the Hugo et al. anti-PD-1 cohort.

  • 83.4 vs 81.7Qwen 3.8 27B, open weights, with the Knowledge Plane, against Claude Opus 5 alone. Plan quality, 0 to 100.
  • Up to 80%lower cost per plan: GLM 5.3 with the plane, against Claude Opus 5 alone.
  • 0invented references in any plan made with the Knowledge Plane.

Inflexa planning benchmark, 54 bulk RNA-seq planning tasks, 14 Sep 2026. Single run, scored by a blind AI judge.

Benchmark

An open model with the plane outplanned a frontier model alone.

The lower-cost and open models gained the most. Qwen 3.8 27B is open weights, so a team can host it on its own GPUs and keep every prompt inside its network.

Grounding lifted every model we tested.

Plan quality, 0 to 100

Plan quality, 0 to 100. The frontier model alone, Claude Opus 5, scored 81.7.
ModelAloneWith the planeCost per plan
Claude Opus 581.790.9$0.72 alone
GLM 5.369.889.2$0.14 with the plane
Qwen 3.8 27B62.983.4$0.09 with the plane
Cost per plan

$0.72

Claude Opus 5 alone

$0.14

GLM 5.3 with the plane

up to 80% lower cost per plan

with a model you choose on your machine

Qwen 3.8 27B with the plane: $0.09 per plan

0 invented references in any plan made with the plane

Inflexa planning benchmark, 54 bulk RNA-seq planning tasks, 14 Sep 2026. Single run, scored by a blind AI judge.

Knowledge PlanePreviewIncluded with every paid plan

Grounded in published methods.

The Knowledge Plane holds rules for when each method applies, drawn from the literature, with tested script templates. Inflexa matches your study design to those rules before it writes a plan, then checks the plan against them.

  1. Your design goes out as counts and categories. A sample row, a file path or an identifier never does.
  2. The plane returns the procedure, each step tied to a rule and its sources.
  3. Standard steps run from tested templates, and the rule behind each one lands in the signed record.

How the Knowledge Plane works

Knowledge Plane

Common practice

When a TPM, FPKM or log table has a known batch the design can hold, put the batch in the limma design. The test never takes adjusted values.

Consensus

When the values are TPM, FPKM or RPKM, use limma-trend on the log values. DESeq2, edgeR and voom need raw counts.

Back on your machine

  1. 1Log transform, log2(FPKM + 1)template
  2. 2PCA by site and strandednesstemplate
  3. 3limma-trend, sex, age, strandednesstemplate
  4. 4fgsea on hallmark setstemplate

Plan checked: no rule broken

  1. Field: Data (bulk RNA-seq, FPKM)
  2. Field: Groups (two-group contrast)
  3. Field: Replicates (14 vs 12)
  4. Field: Batch (batch: site, strandedness)
  5. Field: Covariates (covariates: sex, age)
  6. Field: Pairing (pairing: none)
  7. Rule: Log transform
  8. Rule: Sample structure
  9. Rule: Batch in the design
  10. Rule: Covariates
  11. Rule: FPKM input
  12. Rule: Ranked enrichment
  13. Method: PCA
  14. Method: limma-trend
  15. Method: fgsea
  16. Template: log-scale QC
  17. Template: limma-trend two-group
  18. Template: fgsea preranked
  19. Source: Zhao 2020
  20. Source: Conesa 2016
  21. Source: Nygaard 2016
  22. Source: Law 2014

Your models, your network

Your model, on your GPUs, inside your network.

Pick the model per analysis: Claude or GPT through their APIs, or any OpenAI-compatible endpoint, including open weights on your own cluster. With the Knowledge Plane an open model outplanned a frontier one, so the GPUs you already own can carry the work.

  • Run your own model and no prompt reaches a model vendor.
  • The Knowledge Plane sees the shape of your study as counts and categories. Sample rows, file paths and identifiers stay with you.
  • The model behind every step goes into the signed record by name.

Model for this analysis

Ask a follow-up…Qwen 3.8 27B

Hosted Inflexa Cloud runs Claude. On Enterprise, deployed in your own cloud account, you choose the model.

What crosses the line

Your machine
  • Raw data and sample tables
  • Intermediate files
  • Results, figures and reports
  • The signed record

Your model provider

prompts and excerpts, to the one you chose

Public sources

the lookups and web searches it makes for you

Knowledge Plane

counts and categories, paid plans

The Knowledge Plane never receives sample rows, file paths or identifiers.

Point Inflexa at a model you host and the prompts stay on your network too.

How it works

The agent does the running. You do the deciding.

  1. 01

    Ask

    Describe the biology in plain language. Inflexa reads your data profile and the literature before it proposes anything.

    You decide the question

  2. 02

    Approve

    You get a plan with every method and parameter laid out. Nothing runs until you approve it.

    You decide the plan

  3. 03

    Run

    Each step runs in a sandbox, and independent steps run side by side. Change course at any point.

    You decide when to change course

  4. 04

    Defend

    Results arrive with the record of how they were made. Ask any output where it came from.

    You decide what the biology means

See each step in the workspace

Products

One engine, three ways to run it.

The planner, the sandbox, the signed record, the skill packs and the database connections are the same in every product. Choose where the work runs and how you like to work.

Not sure where to start?

Your data needs to stay
and you work

  • On your machine

    Inflexa CLI

    Open source, in your terminal.

    • Apache 2.0, no account, no meter
    • Any OpenAI-compatible endpoint, including one you host
    • Every step in a sandbox on your machine
    • A signed record you can verify offline

    Free

    curl -fsSL https://inflexa.ai/install.sh | bash
  • On your machine

    Inflexa ProNew

    A full workspace, served from your machine.

    • Browser workspace with a local server
    • The Knowledge Plane, in preview
    • Chat, plan, runs and files side by side
    • Any model, including one you host

    Pricing on request

    Available now. Per seat, per year.

  • In the cloud

    Inflexa Cloud

    Your team's workspace, with compute.

    • Shared projects, with access granted per project
    • Compute that scales as needed
    • Spend caps and alerts by project
    • Runs keep going after you close the browser
    • The Knowledge Plane on Biotech and Enterprise

    Personal free · Biotech · Enterprise

Compared

How Inflexa compares.

Many tools now write and run an analysis for you. The difference is whose model does the thinking, where it runs and who can read the code.

What it doesA chat assistantAn AI scientist from a model providerInflexa
Runs the analysis for youSome do, on the vendor's serversYes, in a sandboxYes, in a sandbox on your machine, your cloud or ours
ModelsThe vendor'sThe vendor'sAny, including open weights you host
Model on your own GPUsNoNoYes, behind any OpenAI-compatible endpoint
Prompts stay on your networkNoNoYes, with a model you host
SourceClosedClosedApache 2.0 engine

Inflexa adds two things to the table. The Knowledge Plane checks each plan against published rules for when a method applies. And every result carries a signed record that anyone can verify offline with an open-source verifier.

See the full comparison

Proof

Real analyses on public data, with the record attached.

Case study · PROTECT cohort

From confounded cohort to druggable hypotheses.

A run on GSE109142, 206 patients: three mechanism-anchored hypotheses, one disqualification the direction analysis forced, and CXCL13 as the candidate the evidence kept returning to.

Read the case study
Translational priority rankingPROTECT cohort · n=206

Once PK, safety and clinical readiness are scored, rosiglitazone falls from the top biology score and sodium butyrate rises to first.

ScorePioglitazoneTofacitinibRosiglitazoneUpadacitinibMesalamine (SoC)Sodium butyrate00.20.40.60.81Biology scoreAfter PK and safety

Questions about Inflexa.

Start with the question you already have.

Install the CLI and pay nobody but your model provider, open a free Cloud account, or talk to us about Pro for your team.