GaiaLab turns biological context into source-linked, testable research hypotheses with visible uncertainty and reproducible records. Plans scale the research workflow—not unsupported claims about model superiority.
Public exploration remains easy to try. Paid plans add durable research workflow, exports and programmatic access. Institutional teams begin with a bounded validation pilot before broader deployment.
Evaluate the basic GaiaLab workflow before committing to a paid plan.
For individual researchers who need a repeatable evidence workflow rather than a one-off answer.
For advanced users who need higher-throughput, programmatic, and audit-oriented workflows.
| Runtime contract | Free | Researcher | Pro | Institutional pilot |
|---|---|---|---|---|
| Analyses per day | 10 | 50 | 200 | Negotiated |
| Maximum genes per analysis | 50 | 50 | 50 | 50 |
Rather than asking a scientific organization to buy a broad enterprise promise, GaiaLab begins with a bounded 4–6 week evaluation against the team's own review process and predefined success criteria.
Review the Validation Pilot →Initial validation pilots should use public, published, synthetic, or otherwise non-confidential inputs. GaiaLab does not currently represent the public workflow as ready for PHI, patient-identifiable data, regulated clinical data, trade-secret datasets, or confidential institutional information. Confidential-data use requires separately agreed terms and verified enterprise controls.
Plans are described by buyer value and workflow. Internal model or agent configurations may change and are not represented as evidence of scientific superiority.
Evidence availability depends on source health, credentials, query context, licensing and deployment status. GaiaLab should report provenance and availability rather than market a fixed count as a scientific guarantee.
Calibration, evaluation and workflow updates are versioned. GaiaLab does not claim that every analysis automatically makes the scientific system more accurate.
Eligible prediction and research records are hashed into Merkle leaves and can be batched through OpenTimestamps. A record is described as Bitcoin-confirmed only after its stored receipt is independently verified; /verify exposes the exact proof status and boundary. Confirmed exact-name forward hits so far, reported honestly: see the live ledger at /new-trials.