Sold by: SandboxAQ
Deployed on AWS
A high-fidelity, spin-aware machine learning interatomic potential family that delivers DFT-level adsorption energy accuracy for industrially relevant catalysts at up to 20,000x the speed of physics-based calculations.
Overview
AQCat is a family of spin-aware machine learning interatomic potentials (MLIPs) built for heterogeneous catalysis screening and chemical discovery. AQCat supports both direct model inference for structural energies and forces and an adsorption workflow for estimating minimum adsorption energies across candidate catalyst surfaces. It is designed to deliver DFT-level adsorption energy accuracy at up to 20,000x the speed of physics-based calculations. Developed for materials scientists, catalyst researchers, computational chemists, and R&D teams in industries including chemicals, energy, oil and gas, agriculture, and pharmaceuticals, it enables users to evaluate catalyst candidates faster while keeping proprietary structures and workflows inside their own environment.
The AQCat25-EV2 model is trained on the AQCat25 dataset with 13.5 million high-fidelity DFT calculations. For organizations seeking a machine learning force field (MLFF) for adsorption energy prediction, transition-state exploration, and virtual screening, AQCat is designed to reduce the tradeoff between quantum-chemistry accuracy and computational throughput.
Extended customer support including direct access to an engineer (8am-8pm ET) is available for this product for a yearly fee. Please contact us at support.aisim@sandboxaq.com for more information.
Note: We use AWS's Standard Contract, except that we do not carry the professional liability insurance referenced in the standard contract. Ordering through AWS marketplace constitutes an acceptance of that.
Highlights
- In-Silico Virtual Screening: Pre-screen massive, high-risk chemical design spaces and novel bulk compounds in days instead of months to optimize R&D budgets before physical synthesis.
- Isolated, Secure Workloads: Run highly proprietary molecular structures and catalyst designs with absolute data privacy. Because the model deploys via Amazon SageMaker inside your own isolated AWS tenant, your input data, inference payloads, and chemical structures are never exposed to external networks, never shared with SandboxAQ, and never used for retraining.
- Unlocking Sustainable Energy and Chemicals: Accelerate the discovery of optimal active sites and intermediates for green hydrogen production (water splitting), carbon capture utilization (CO2-to-fuels), fuel cells, sustainable fertilizer production, and methane-to-methanol conversion.
Details
Sold by
Categories
Delivery method
Latest version
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
Pricing is based on actual usage, with charges varying according to how much you consume. Subscriptions have no end date and may be canceled any time.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimension | Description | Cost |
|---|---|---|
ml.g4dn.xlarge Inference (Batch) Recommended | Model inference on the ml.g4dn.xlarge instance type, batch mode | $999.00/host/hour |
ml.g5.xlarge Inference (Batch) | Model inference on the ml.g5.xlarge instance type, batch mode | $999.00/host/hour |
inference.count.m.i.c Inference Pricing | inference.count.m.i.c Inference Pricing | $0.002/request |
Dimensions summary
You pay based on usage, with no upfront commitment. Two dimensions bill by host hour for running model inference in batch mode. One uses the ml.g5.xlarge instance type, the other uses the ml.g4dn.xlarge instance type. Your choice of instance affects the hourly rate, so you select the one that fits your workload. A third dimension bills per inference request, charging you for each prediction the model returns. Together, these let you pay either for the time an instance runs or for the number of requests you make.
Top-of-mind questions for buyers
A host hour meters the running time of one inference instance in batch mode. You are charged for each hour the ml.g5.xlarge or ml.g4dn.xlarge instance runs. Your choice of instance sets the hourly rate, so time running and instance type together drive your batch cost.
Each request is one prediction the model returns, such as an adsorption energy or binding configuration result for a catalyst system. You are charged per request, so cost scales with how many predictions you run rather than with how long any instance stays active.
The batch host-hour dimensions and the per-request dimension bill independently. Host-hour charges accrue while an instance runs in batch mode. Request charges accrue per prediction returned. You pay only for the dimensions you use, so time-based and request-based costs appear separately rather than stacking on the same workload.
www.sandboxaq.com
Helpful?
Vendor refund policy
All charges and subscription fees for this SaaS offering are final and non-refundable. You may cancel your subscription at any time to prevent future billing; however, your cancellation will take effect at the end of your current billing cycle. We do not provide prorated refunds or credits for any partial months, unused time, or downgraded plans.
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
Amazon SageMaker model
An Amazon SageMaker model package is a pre-trained machine learning model ready to use without additional training. Use the model package to create a model on Amazon SageMaker for real-time inference or batch processing. Amazon SageMaker is a fully managed platform for building, training, and deploying machine learning models at scale.
Deploy the model on Amazon SageMaker AI using the following options:
Version release notes
Constraints are honoured when relax is set — an unconstrained slab relax would move the held atoms and return an energy not comparable to the workflow path. They remain ignored for a single-point pass, where they cannot affect the result.
Additional details
Inputs
- Summary
There are two modes that the model can be run in. More details and input samples can be found in the docs here .
For mode: "mlp" (default, docs here ), the model expects application/json input with an instances array of one or more atomic structures, where each structure is provided as a JSON-serialized ASE Atoms object. Multiple structures can be included in the same instances array to create a mini-batch within a single request.
For mode: “min-adsorption-energy-workflow” (docs here ): The model expects application/json input with an instances array of one or more screening requests, where each instance represents a bulk/facet/adsorbate combination for catalysis screening. For both modes, the real-time endpoint is the primary and recommended invocation path. Batch transform is possible and uses the same JSON schema.
- Input MIME type
- application/json, text/csv
https://github.com/sandbox-quantum/marketplace-public-docs/blob/main/aqcat/aws-marketplace/mlp-model/sample_data/input_realtime_sample.json
https://github.com/sandbox-quantum/marketplace-public-docs/blob/main/aqcat/aws-marketplace/adsorption-workflow/sample_data/input_batch_sample.json
Resources
Vendor resources
Support
Vendor support
Support resources
For all support inquiries - including model deployment assistance, inference troubleshooting, accuracy questions, billing issues or more generally getting started - contact the SandboxAQ AQCat team at support.aisim@sandboxaq.com
Please consult our product's documentation listed under the Resources section before reaching out.
Refund policy
All charges and subscription fees for this SaaS offering are final and non-refundable. You may cancel your subscription at any time to prevent future billing; however, your cancellation will take effect at the end of your current billing cycle. We do not provide prorated refunds or credits for any partial months, unused time, or downgraded plans.
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
SA On Demand: Get free technical advice or validation from an experienced, certified Solutions Architect. Get guidance or help deploying and configuring new or existing AWS services.
The Data Cleaning & Preparation for LLM Training Assessment helps organizations evaluate and prepare high-quality, secure datasets for GenAI, fine-tuning, and RAG workloads on AWS. Led by senior AI and AWS architects, this assessment reviews data quality, governance, and preprocessing readiness to deliver a clear roadmap aligned with AWS best practices and services in mind.
AI innovation creates four critical risks for startups: it can consume valuable runway, disrupt the core roadmap, delay time to market, and ultimately result in the wrong capability being built.
The Agentic AI Catalyst Program by LorinCloud is a specialist prototype and Proof of Concept (PoC) delivery program designed to reduce these risks. It enables startups to validate one commercially important AI capability quickly through a milestone-led, right-sized investment without the cost and complexity of premature and high-cost overengineering.
The program helps you turn an innovative idea into a working early-stage solution, validate its potential, create a customer-facing demo, or showcase AI capabilities to investors.
From voicebots and recommendation systems to sales agents, contact center AI, and video analytics, we provide world-class AI engineering expertise to build and prove your next AI capability in the cloud with Bedrock.
CATALYST embeds the AWS AI-Driven Development Lifecycle in your engineering org — Kiro + Amazon Q Developer configured, teams trained, bolt cycles supervised, and a Center of Excellence established. AI-DLC adoption in 90 days.
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.