Frequently Asked Questions
25 straight answers about what AIHub costs, what a free account unlocks, how to contribute, how profiles are verified, and how the platform stays independent of the vendors it covers.
Access & cost
How much does it cost to use AIHub?
Nothing. AIHub is free to use and there is no paid tier, subscription, or premium plan. Browsing the tools directory, company profiles, comparisons, learning resources, and news requires no account at all. A free account unlocks the full contents of case studies, compliance framework guides, and the consulting toolkit playbooks.
What requires a free account, and why?
Three sections require a free sign-in to read in full: the 73 case studies, the 35 compliance framework guides, and the 37 consulting toolkit playbooks. Their index pages — titles, summaries, and what each covers — are public. The sign-in exists to understand who the material serves, not to charge for it.
Does AIHub sell consulting engagements or implementation services?
No. AIHub is a knowledge platform, not a consultancy. It publishes playbooks, templates, and checklists you run yourself. There is no billable engagement, no delivery team, and nothing to purchase.
Contributing & listings
How do I create and submit a case study to AIHub?
Submit it through the contribute page. A case study needs the company and industry, a clear problem statement, the AI solution used, measurable outcomes with real metrics, and a public source or reference that verifies them. Submissions go to the editorial team for review. Only publicly available information is accepted — no proprietary or confidential material.
How do I list my company in AIHub's vendor profiles, and what are the requirements?
Submit it through the contribute page with the company name, HQ and founding year, focus area and stage (public, private, or research lab), key products and models, funding if publicly disclosed, and a brief history. Entries are editorially reviewed against public sources. Listing is free and cannot be bought — there is no paid placement, sponsored listing, or advertising on AIHub, which is what keeps the directory's 83 company profiles independent.
Can I pay to have my tool featured or ranked higher?
No. There is no paid placement, sponsorship, or advertising anywhere in the directory. Rankings and inclusion are editorial decisions based on public information, and AIHub is not affiliated with any listed vendor.
Accuracy & maintenance
How accurate and up-to-date are the company and vendor profiles?
Profiles are hand-verified against public sources — official documentation, published financials, research papers, and company announcements — rather than vendor-supplied copy, and are updated daily. AIHub is independent and not affiliated with any listed company. Because the AI market moves quickly, always confirm pricing and capability details against the vendor's own documentation before making a purchasing decision.
Are all case studies backed by a public source link?
Most carry direct reference links to company disclosures, peer-reviewed research, or official documentation, and those are the ones to rely on if you are citing a figure. A minority document well-reported deployments without a link attached yet; adding sources to those is ongoing editorial work. Figures that were never publicly disclosed are marked as such rather than estimated, so an unmarked number should always be checked against its source before you use it in a business case.
How does AIHub keep its regulatory and compliance information current?
The 35 frameworks across 25 jurisdictions are reviewed and updated as regulations change, with each entry recording its status, enforcing authority, key requirements, timeline, and penalties. Regulation moves faster than any summary, so the guides are educational reference rather than legal advice.
Using the platform
How do I evaluate which tools are production-ready versus research-only?
Every entry in the 147-tool directory records its pricing model, platforms, integrations, underlying models, and documented use cases. Signals that a tool is production-ready include an enterprise or paid tier, published integrations with systems you already run, a stated support model, and named production deployments. Research-oriented tools typically show none of these. The consulting toolkit's vendor evaluation scorecard turns this into a weighted, repeatable comparison rather than a judgement call.
How do I find tools for a specific use case?
The directory is organised into 13 categories — including language models, code assistance, image and video generation, voice and audio, infrastructure, frameworks, APIs, search and research, and autonomous agents — and is filterable by category, pricing, and use case. Once you have a shortlist, the comparison pages put two tools side by side on pricing, features, integrations, and best-fit use cases.
How do I use AIHub's case studies to build a business case for AI investment?
Each of the 73 case studies documents the business context, what was actually built, the measured outcome, investment estimate where disclosed, and lessons learned — including what went wrong. Find the studies closest to your industry and problem shape, use them to sanity-check the size of the prize and the time to value, then build the numbers with the toolkit's AI ROI Measurement Framework so your assumptions are explicit and reviewable. Figures that are not publicly disclosed are marked as such rather than estimated.
Which learning resource should a product manager start with?
The 108 learning resources are tagged beginner, intermediate, and advanced and filterable by type — courses, hands-on code tutorials, video playlists, certifications, and books. Product managers generally get furthest starting with a beginner-level conceptual course to build accurate intuition about what models can and cannot do, before touching anything hands-on.
Consulting toolkit
What is the difference between AIHub's consulting toolkit and hiring an AI consultancy?
They solve different problems. The toolkit gives you 37 playbooks with fill-in checklists and templates covering assessment, pilot, scale, and governance — the structure and questions a good engagement would bring, at no cost, run by your own team on your own timeline. A consultancy additionally brings people, delivery capacity, accountability for outcomes, and pattern-matching from engagements you cannot see. The toolkit is a substitute for the framework, not for the hands.
How long does it take to implement a playbook in a mid-size company?
It depends on the playbook and how much of the work is organisational rather than technical. Assessment playbooks are typically days to a few weeks. Pilot playbooks run over the pilot itself — commonly six to eight weeks for a first cohort. Scaling and governance playbooks are ongoing programmes rather than one-off exercises. The binding constraint is almost never the template; it is decision-making, data readiness, and change capacity.
Does AIHub provide AI impact assessment templates?
Yes. The toolkit includes fill-in templates for AI readiness assessment, enterprise maturity scoring, data estate and permission auditing, use-case prioritisation by value against feasibility, model risk management, and responsible AI and bias testing. Whether you are formally required to conduct an impact assessment depends on your jurisdiction and risk tier — under the EU AI Act, high-risk systems carry specific obligations — so confirm the requirement with qualified counsel and use the templates to do the work.
My AI project missed its ROI target — what does AIHub offer?
Start with the AI ROI Measurement Framework to establish whether the target was measurable and baselined in the first place, since unmeasurable targets are the most common cause of apparent failure. Then work through the Change Management and workforce adoption playbooks: value usually stalls at adoption and workflow redesign rather than at model quality. The case studies document what went wrong in real deployments, which is often more useful than what went right.
Compliance & responsible AI
Do I need legal counsel in addition to AIHub's compliance guides?
Yes. The compliance guides are educational reference — scope, obligations, timelines, and penalties explained in plain language — and are explicitly not legal advice. Any binding decision about whether an obligation applies to your organisation, particularly conformity assessment under the EU AI Act, needs qualified legal counsel in the relevant jurisdiction.
Do I need to retrain my models to follow AIHub's responsible AI recommendations?
Usually not. Most responsible-AI guidance concerns process rather than model weights: documenting intended and out-of-scope use, testing for bias against defined thresholds, adding human oversight proportional to impact, logging decisions for audit, and monitoring for drift in production. Retraining becomes necessary only when testing shows the model itself produces unacceptable outcomes that guardrails cannot mitigate — which is a finding from evaluation, not a starting assumption.
Which compliance framework applies to processing health data?
Typically several at once rather than one. In the EU, health data is special-category data under the GDPR, and an AI system making decisions about care or eligibility is likely high-risk under the EU AI Act. Medical-device software carries its own regime. Sector and jurisdiction determine the actual set, so use the framework guides to identify candidates and confirm applicability with counsel.
My model produces biased or unfair results — what should I do?
Work through the Responsible AI and Bias Testing playbook. It covers defining fairness metrics appropriate to the decision being made, testing outcomes across affected groups, setting thresholds that trigger investigation, and deciding between mitigation in the data, the model, or the decision process around it. Bias is measured against a definition you choose deliberately — the first step is agreeing which definition applies to your use case.
Independence & neutrality
Does AIHub recommend TensorFlow or PyTorch for production ML?
AIHub is vendor-neutral and does not issue recommendations of that kind. Both are listed in the frameworks category with their characteristics documented so you can judge against your own constraints — existing team skills, deployment target, serving infrastructure, and ecosystem fit usually decide this far more than any general ranking would.
How does AIHub compare to other AI directories for enterprise procurement?
They serve different purposes. Model hubs and repositories are built for practitioners to find, run, and host models and datasets. AIHub is built for evaluation and governance: side-by-side tool comparisons, 83 vendor profiles, 73 deployment case studies with measured outcomes, 35 compliance frameworks, and implementation playbooks. For procurement, the useful distinction is that AIHub is independent of the vendors it profiles and takes no paid placement, so nothing in the directory is bought.
Community & collaboration
How can a researcher or PhD student find collaborators through AIHub?
Through AI Hub Swiss's practitioner community, which is free to join and organised into four standing working groups: AI Governance and the EU AI Act, Frontier Research, Enterprise Adoption, and AI Literacy and Talent. The groups deliberately mix people who set AI policy with people who ship AI systems, so the most direct route is to join the working group closest to your research area rather than approaching researchers or industry separately.
Are there AIHub events or meetups?
Yes. The community runs roundtables and practitioner workshops, both online and in person in Switzerland, listed on the network page for members. Membership is free and open to practitioners worldwide.
Still have a question?
Submit a correction, a missing tool, company, or case study through the contribute page, or email info@aihubswiss.org.