Straight answers, before you sign up.
Exactly what we monitor (the AI models behind your apps — LLMs, image, video, and audio), what counts as a workspace, how the plans differ, what a synthetic baseline is, and why we never need your client's data. Current pricing lives on the plans page.
What exactly does AI Stack Watch monitor?
The AI models your software runs on — and the providers behind them. That includes the large language models (LLMs) behind chat assistants and agents, the embedding models behind RAG and search, and the image, video, audio, and speech models you call from providers like OpenAI, Anthropic, Google, Meta, Mistral, AWS, and Azure. AI Stack Watch tracks each model's pricing, capabilities, and lifecycle — launches, deprecations, and retirements. It is not uptime monitoring or analytics for your own servers; it watches the model providers you depend on — the part of your stack you don't control.
What counts as a workspace?
A workspace is one monitored AI stack — typically one client, product, or pipeline. You group the providers and models that pipeline depends on, and AI Stack Watch scopes every alert and brief to that workspace, so you always know whose stack a change touches. Your plan determines how many workspaces you can run.
How do the Free, Solo, Team, and Agency plans differ?
Free is the public dashboard — browse the live AI landscape, no card required. Solo is for an individual watching a single stack of their own. Team is for a small shop watching a handful of client stacks. Agency is for consultancies and fractional CTOs running a portfolio of clients, and adds the white-label monthly brief and priority alerts. Most customers who manage more than one stack start at Team or Agency. Current pricing is on the plans page.
What is an approved validation baseline?
It's a synthetic, customer-approved reference for a workspace — built from synthetic cases, never your client's real prompts or data. You approve it once, and every future model change is evaluated against it, so "should we switch?" has an evidence-based answer instead of a guess. You can edit and re-approve it as the workspace evolves.
What is a validation impact check?
When a replacement candidate actually matters, AI Stack Watch can compare your current model and selected candidates against the approved synthetic baseline and report how each did against the workspace's must-pass checks. It's decision-support — it surfaces which candidates fit the stated constraints, not a verdict on which model is "best."
What happens if a model can't be run?
We tell you plainly. If a provider isn't callable in a given run, that candidate is reported as not run — with the reason — rather than guessed at or scored anyway. An honest gap is more useful than a confident-looking result that isn't real.
Do you need our clients' data or API keys?
No. Validation baselines are synthetic, and AI Stack Watch monitors the providers' own public pages — not your client's systems. We don't need your client's real prompts, private records, files, or API keys to produce briefs or run checks.
What does "decision-support, not decision-making" mean?
AI Stack Watch tells you what changed, why a workspace may be exposed, and what to evaluate against your approved baseline. It never tells you what to do, declares a "best" model, or makes a migration decision for you. The call always stays with you.
Can I white-label the briefs?
Yes. The monthly brief is designed to be sent under your brand. On the Agency tier the "Powered by Enginerds" line is configurable, so you can present it as a fully white-label deliverable to your clients. See the sample artifacts for what that looks like.
Can I cancel anytime?
Yes — plans are self-serve and you can cancel anytime, no phone call required. Billing is handled securely through Stripe.
Is the public AI dashboard really free?
Yes. The live change feed, model directory, provider pages, and cross-provider insights are free for everyone, with every value linked to its source. The paid plans add per-client workspaces, approved baselines, alerts, and white-label briefs.
Ready to look?
Running a portfolio of clients? See how agencies use it.
Curious how we keep the data accurate? Read our methodology.