Your model designed it.
Our autonomous lab tests it.
We operate a compact, fully autonomous R&D lab in Hyderabad. It is running now. We are opening selected programs to academic and industry teams while raising capital to expand its instruments, assays, and throughput.
Biology is becoming programmable
Models can now generate proteins, sequences, constructs, and experimental plans faster than laboratories can test them. The scarce resource is no longer a plausible design. It is trustworthy physical evidence.
Programmable biology needs an execution layer: software translates intent into lab operations, automation runs the experiment, and structured results flow back into the model. Our Hyderabad lab is the first compact implementation of that loop.
Design → execute → measure → learn. The product is not a robot or a report. It is a repeatable experimental loop.
From computational design to experimental evidence
Initial programs are reviewed with our team so we only accept work the current cell can execute reliably. Standardized workflows will move into transparent, self-serve ordering as the public catalog expands.
Start with a working cell. Scale into a network.
Automation already running
The core execution loop is operational in our Hyderabad R&D setup. Funding expands breadth and throughput rather than financing a first demonstration.
Price parity as a baseline
Our launch targets match published cloud-lab prices. The India cost base is intended to preserve margin while keeping experiments accessible.
Model-ready output
Each standardized assay is designed to return structured results, raw files, QC context, and execution provenance.
Clinical context is proposed
We are pursuing, but have not signed, an Apollo partnership for consented clinical and biobank access. It is an expansion path, not a current asset.
One lab, two access models
Academic R&D
Small, grant-sized experimental batches for teams whose computational output is growing faster than their bench capacity.
Industry R&D
Pilot programs now, followed by reserved capacity and API-driven design–build–test–learn loops as the fleet scales.
From design to data
The lab is operational. Access is deliberately reviewed while the catalog is small; standardized assays move to self-serve ordering as capacity expands.
Submit an expression of interest
Tell us the assay class, sample type, approximate batch size, and desired turnaround. Do not send confidential sequences or patient data through the public form.
Feasibility review and quote
We map the request to the current Hyderabad cell, identify any method-development work, and return scope, price, inputs, and an estimated completion window.
Secure technical intake
Accepted programs receive a secure channel for sequences, protocols, material-transfer details, and acceptance criteria.
Autonomous execution
LabOS schedules the workflow, drives the connected instruments, monitors controls, and records the parameters and materials used in the run.
Structured results
Customers receive parsed results, raw instrument files, QC context, and an execution record suitable for analysis or the next model-training cycle.
Pricing targets benchmarked to the market
RNA synthesis + qPCR
Automated synthesis, purification, and quantitative readout for standardized inputs.
Cell-free expression + A280
Expression, affinity purification, yield quantification, controls, and raw data.
E. coli expression + A280
Transformation through expression, purification, yield measurement, and QC.
Thermal-shift characterization
Expression, purification, and standardized thermal unfolding measurements.
Plate-reader assay onboarding
One method-development iteration and qualification run for a compatible assay.
SPR target onboarding
Target qualification for subsequent binder-kinetics programs.
Benchmark checked 31 July 2026 against the public Ginkgo Cloud Lab protocol catalog. Programmable Bio intends to match these published prices at launch; this comparison should be rechecked before publication.
What every standardized run should return
- Typed results — stable fields for the assay class rather than values trapped in prose
- Raw files — instrument-native output retained alongside parsed results
- Execution provenance — instrument, method version, reagent lots, calibration state, and parameters
- QC context — controls, flags, re-run decisions, and failure-mode labels
- Customer control — no reuse for shared datasets without a separate, explicit agreement