|
One agent across
every discipline
Cortexa reads your artifacts, queries verified evidence, runs the analysis, and keeps every claim in the final answer cited — across biomedical, clinical, genomics, chemistry, structures, AI, physics, engineering, materials, energy, and policy work.
Autonomous Research Agent
Ask a complex question in any discipline — Cortexa plans the approach, picks the right sources for the domain, queries them in parallel, cross-references results, and synthesizes a cited brief. Dozens of tool calls, zero babysitting.
Literature & Evidence Synthesis
Queries arXiv, Semantic Scholar, PubMed, bioRxiv, and domain-specific archives simultaneously. Returns ranked evidence with inline citations, not a list of links.
Life-Sciences Depth
Fetches structures from AlphaFold and PDB, runs variant lookups in ClinVar, Open Targets analyses, and visualizes 3D conformations — all inside the conversation.
Trials, Patents & Policy
Parses recruiting trials, patent landscapes, and regulatory guidance. Returns matched records with IDs, assignees, deadlines, and the primary source — ready to act on.
Clinical Evidence Intake
Drop in radiology reports, lab panels, pathology notes, genomics reports, or clinical summaries. Cortexa detects the artifact, extracts what the report actually states, preserves uncertainty and negations, then connects the findings to cited literature, guidelines, trials, labels, and coding systems. Built for research context — not diagnosis or treatment orders.
Runs the Analysis, Not Just the Search
When a question needs a number — power calculations, hazard ratios, regressions, custom plots from a public dataset — Cortexa writes Python and runs it in the Cortexa compute sandbox. NumPy, pandas, SciPy, scikit-learn, statsmodels, and matplotlib are pre-installed; figures and tables come back inline as artifacts. Real computation, fully cited inputs, no pasting code into another tab.
Safety, Grounded in the Right Domain
Cortexa applies the right safety frame per question — research-framed clinical caveats, drug-label constraints, ML reproducibility notes, legal jurisdiction warnings, or nothing at all for general academic work.
Inline structures, charts, formulas
When a question deserves a visual, Cortexa renders it inline — molecular structures from PDB / AlphaFold, lightweight charts, LaTeX formulas, and elevated citation cards next to the prose.
API for Agentic Research
OpenAI-compatible endpoint lets your team drop Cortexa into EHRs, lab notebooks, ML training pipelines, patent-review systems, and internal tooling. Metered billing, zero lock-in.
Cortexa, wherever
you already work.
One research agent, reachable from every surface your team lives on. Same plan, same citations, same usage — no matter which one you invoke it from.
The research canvas
The full Cortexa experience — autonomous research, inline structures and charts, and cited briefs you can print or share.
- Stream tool calls in real time
- Drop in PDFs, papers, notes
- Generate a formal research brief
Cortexa in Slack
@mention Cortexa in any channel, DM it privately, or run /cortexa from the composer. Same agent, same plan, same citations.
- @cortexa in channels & threads
- /cortexa from any composer
- Per-workspace install · SSO-ready
Private connections
Bring your own Postgres, S3, GCS, Bitquery, Google Drive, or Box context into Cortexa through encrypted user-scoped connections.
- Tailored setup forms
- Encrypted credentials
- OAuth-ready document apps
MCP server
Call Cortexa as a tool from Claude Desktop, Cursor, Codex, Claude Code, Windsurf, Zed — any MCP-capable client. Stdio via npx or hosted streamable-HTTP.
- npx -y @yoreai/cortexa-mcp
- Hosted at /mcp for remote clients
- One cortexa_research tool, cited answers
Developer API
OpenAI-compatible endpoint for wiring Cortexa into EHRs, lab notebooks, pipelines, or your own chat UI. Metered, no lock-in.
- Drop-in /v1/chat/completions
- Usage-based pricing
- Scoped API keys per environment
Shipping surfaces is our superpower
Slack went from idea to production in under a week — the Chat SDK under the hood means we can light up another surface fast. Here's what's queued up next.
What Cortexa reads from every session
Fifteen research surfaces, one agent. From biomedical to finance, Cortexa picks the right sources for the question, queries them in parallel, and cites every finding in the final brief.
Pricing
Start free. Upgrade when the research gets serious.
Start free, then step up to Plus, Pro, or Enterprise. Plan management, checkout, and team controls all live inside your Cortexa billing settings after sign-in.
Free
Try the agent with real cited research.
50 credits / month
Start free- No credit card required
- All 2.2K+ verified tools
- Shareable sessions
Plus
For individual researchers who want real headroom without the Pro commitment.
750 credits / month
Choose Plus- Everything in Free
- 15× the Free credit budget
- Overage cap you control
Pro
For regular research work with more credits and overage headroom.
3,000 credits / month
Choose Pro- Higher monthly credit budget
- Overage cap you control
- Heavier files, tools, and connected-data budgets
Enterprise
For teams that need one shared credit pool, controls, and private data access.
35,000 shared credits / month
Team settings- No per-seat pricing to start
- Shared private connections
- Member usage visibility
One credit maps to actual AI work. Teams share a credit pool instead of paying per seat.
Free pauses after the included budget. Paid plans can auto-continue at $0.012/credit up to a cap you set.
Every plan — Free, Plus, Pro, and Enterprise — reaches the same 2.2K+ verified tools, files, connections, API, and MCP surface. Credits control usage.
Need details before choosing?
Compare plans after sign-in, including checkout, credits, and team controls.
Ready to accelerate
your research?
Open the agent, ask anything worth knowing, and get a cited brief you can revise and share.