← All companies

Science / Company profile

SandboxAQ

What SandboxAQ does

SandboxAQ is an independent B2B company applying “Large Quantitative Models” (LQMs)—physics, chemistry, and scientific-equation grounded models—alongside AI and (where relevant) quantum-tech techniques to produce decision-grade outputs for high-stakes, real-world domains. The company traces its origins to Alphabet and says it has operated as a standalone organization since 2022.

Across its portfolio, SandboxAQ’s offering is organized around three enterprise problem areas: (1) life sciences and chemicals/materials discovery, where it positions models and simulation workflows as a way to accelerate R&D cycles; (2) cryptography and AI security posture management, where it focuses on inventory, governance, and remediation for cryptographic assets; and (3) navigation and sensing, where it develops GPS-denied positioning approaches and related sensing-based diagnostics.

Business model and customers: SandboxAQ sells enterprise solutions and software/model access (for example via cloud/marketplaces and interfaces intended to reduce coding burden) to organizations that run mission-critical workflows in regulated or compute-intensive environments—such as life-sciences discovery teams, industrial materials developers, and government/defense or enterprise security teams.

Strategic positioning: In 2026, SandboxAQ emphasized broader distribution of its quantitative models through AI ecosystems and marketplaces (e.g., integrations with Claude via Model Context Protocol, and distribution via Google Cloud’s Marketplace). It also announced large public-sector funding tied to semiconductor materials and chemicals discovery, reinforcing its positioning that its LQM + simulation approach is intended to run at scale for physical-world R&D.

News

Company record

Aug 19, 2026 · Official · SandboxAQSandboxAQ Makes AQCat Generally Available on Claude for High-Throughput Catalyst Screening

SandboxAQ announced general availability of AQCat on Claude via Model Context Protocol (MCP), describing AQCat as spin-aware for magnetic behavior and positioned for fast catalyst screening from natural-language prompts.

Aug 19, 2026 · Official · SandboxAQSandboxAQ Launches AQPotency, Bringing Ultrafast Virtual Screening to Drug Discovery Without a Solved Protein Structure

SandboxAQ announced general availability of AQPotency on Claude via MCP, describing it as a Large Quantitative Model for scoring candidate molecules for potency/target interaction without requiring a solved protein structure.

Jun 29, 2026 · Official · SandboxAQSandboxAQ to Bring Large Quantitative Models to Google Cloud's Marketplace

SandboxAQ announced it would bring its Large Quantitative Models to Google Cloud’s Marketplace to broaden availability for science research use cases.

Jun 23, 2026 · Official · SandboxAQSandboxAQ Launches Virtual Screening Solution for GPCR Drug Discovery, Accelerated by NVIDIA BioNeMo

SandboxAQ announced a GPCR virtual screening solution positioned to predict both binding and whether compounds act to activate or block receptor states, and it said the workflow is accelerated using NVIDIA BioNeMo’s Agent Toolkit.

Jun 22, 2026 · Official · SandboxAQSandboxAQ Supports White House Executive Order as Post-Quantum Cryptography Moves to Implementation for National Security

SandboxAQ released a statement tying the company’s work to the White House’s post-quantum cryptography executive order and described support for the federal PQC transition agenda toward implementation milestones.

Jun 17, 2026 · Official · SandboxAQSandboxAQ Signs Definitive Agreement with the U.S. Department of Commerce for $500 Million CHIPS R&D Award

SandboxAQ announced a definitive agreement for a $500M CHIPS Research & Development award, describing work spanning PFAS-free process chemicals, catalysts, rare-earth-free magnets, and battery systems, and stating Commerce receives a minority, non-voting equity stake.

May 18, 2026 · Official · SandboxAQSandboxAQ Integrates its Quantitative AI Models with Anthropic's Claude via MCP

SandboxAQ announced integration of its LQMs with Claude via MCP, describing the goal of enabling direct connections between an LLM interface and quantitative models for drug discovery and materials science use cases.

Jan 20, 2026 · Official · SandboxAQSandboxAQ Supercharges OpenFold3 with AQAffinity: High-Speed, Structure-Free Drug Potency Prediction

SandboxAQ announced an enhancement to OpenFold3 via AQAffinity, positioning it as a high-speed approach for drug potency prediction that is structure-free.

Show 3 earlier updates
Dec 03, 2025 · Official · SandboxAQSandboxAQ Launches AQtive Guard AI-SPM to Stop the Rapid Spread of Shadow AI Across the Enterprise

SandboxAQ announced AQtive Guard AI-SPM, describing it as AI security posture management designed to discover, analyze, and mitigate AI risk at scale.

Sep 10, 2025 · Official · SandboxAQSandboxAQ Releases AQCat25 Dataset, Accelerating Next-Generation Catalysis and Materials Discovery with AI

SandboxAQ announced AQCat25, a large-scale dataset it said contains 11 million high-fidelity quantum chemistry calculations designed to support catalysis and materials discovery predictions.

Jul 31, 2025 · Official · SandboxAQSandboxAQ Launches AQtive Guard Protect to Combat the Mounting Security Threats Posed By Non-Human Identities

SandboxAQ announced general availability of AQtive Guard Protect, positioning it as a solution for managing and securing non-human identities and cryptographic assets and describing it as built on AQtive Guard Discover.

Source map · 2 recurring channels · 14 references

Still resolving: Newsroom · Official X

Funding

Latest disclosed valuation$5.3B
Tracked capital$2.1B7 sourced rounds
DateRoundRaisedValuationLead / investorsEvidence
Jun 18, 2026Series F
Not attributed
Jun 17, 2026Government award (CHIPS R&D award; non-dilutive stated by framing in release)$500M

What it builds

AQCat

SandboxAQ’s AQCat is a catalysis-focused large quantitative model intended for catalyst screening, described by the company as “spin-aware” for magnetic behavior and designed to run at scale from natural-language prompts when made available through Claude via Model Context Protocol (MCP).

source ↗
AQPotency

SandboxAQ’s AQPotency is a Large Quantitative Model for drug discovery potency/interaction scoring that the company says can rank candidate molecules without requiring a solved protein structure, and that is available through Claude via MCP.

source ↗
AQtive Guard

AQtive Guard is SandboxAQ’s cryptography management/security solution that the company describes as enabling encryption remediation and automated/supervised enforcement while continuously optimizing cryptography across an organization.

source ↗
AQNav

AQNav is SandboxAQ’s navigation solution designed to support assured positioning when GPS is jammed, spoofed, or denied, combining quantum sensing, AI, and Earth’s magnetic field for location estimation.

source ↗
AQMed (CardiAQ)

AQMed develops next-gen medical diagnostics using AI and advanced sensing; its first product, CardiAQ, is described as a magnetocardiography (MCG) investigational device that aims for faster, radiation-free cardiovascular diagnostics via magnetic signal capture and analysis.

source ↗
AQChemSim

AQChemSim is SandboxAQ’s chemistry and materials simulation platform positioned as a way to simulate the physical world before entering the lab, using LQMs trained on physics/chemistry/materials science to generate high-fidelity predictions.

source ↗

Milestones & partnerships

Aug 19, 2026
AQCat made generally available on Claude via MCP

SandboxAQ announced general availability of AQCat on Claude Science through Model Context Protocol (MCP).

Anthropic (Claude)
Aug 19, 2026
AQPotency made generally available on Claude via MCP

SandboxAQ announced general availability of AQPotency on Claude via MCP for structure-free virtual screening / potency scoring.

Anthropic (Claude)
Jun 23, 2026
GPCR virtual screening solution launched with NVIDIA BioNeMo acceleration

SandboxAQ launched a GPCR virtual screening workflow that predicts binding and functional activity (agonist/antagonist/inverse agonist framing) and said it is accelerated by NVIDIA BioNeMo’s Agent Toolkit.

NVIDIA
Jun 17, 2026
U.S. Department of Commerce CHIPS R&D award agreement (re: materials/chemicals programs)

SandboxAQ announced a definitive agreement for a $500M CHIPS Research & Development award, covering PFAS-free process chemicals, catalysts, rare earth-free magnets, and battery systems; it also stated Commerce receives a minority, non-voting equity stake.

U.S. Department of Commerce (CHIPS Research and Development Office)
May 18, 2026
LQM integration with Anthropic Claude via MCP

SandboxAQ announced LQM integration with Claude through Model Context Protocol so that an LLM can serve as a natural language interface to the quantitative model outputs.

Anthropic
Sep 10, 2025
AQCat25 dataset launch

SandboxAQ announced AQCat25 with 11 million high-fidelity quantum chemistry calculations to support catalysis discovery and materials discovery predictions.