Skip to content
WebmasterID

Demos

Guided product demos

Follow source-backed workflows that show how use cases, shortlists, comparisons, evidence briefs, sources, and freshness fit together. These demos are navigation examples, not model recommendations.

Pick a demo

Three guided workflows

Each demo loads the same verified data layer the rest of the site uses — the demo only changes which fields you walk through first.

  • Guided demo

    Long-context analysis

    Walk a long-context (≥200k-token) workload: pick the use case, narrow a shortlist by verified context window, compare the candidates side by side, and export an evidence brief. Pricing tier references for prompts >200k are flagged on the comparison; the brief surfaces every data gap explicitly.

    Models
    4 in walkthrough
    Fields inspected
    context window · max output tokens · modality channels…
    Start walkthrough →Jump to evidence brief

    Navigation example, not a model recommendation.

  • Guided demo

    Hosted inference

    Walk a hosted-inference workflow where the model creator does not run a paid first-party API. Pick the use case, narrow the shortlist to models with verified hosted availability, then compare hosted pricing references side by side. The brief separates the hosting platform (billing provider) from the model creator at every step.

    Models
    2 in walkthrough
    Fields inspected
    hosted availability record (host × model) · hosted model ID · billing provider…
    Start walkthrough →Jump to evidence brief

    Navigation example, not a model recommendation.

  • Guided demo

    Governance review

    Walk an internal AI-inventory + source-backed due-diligence flow. Pick the use case, narrow a shortlist with verified citations, compare verification state side by side, then generate an evidence brief that lists every data gap and stale citation explicitly. The reverification queue is the natural next stop.

    Models
    4 in walkthrough
    Fields inspected
    verification status · lifecycle status · source citations with retrievedAt…
    Start walkthrough →Jump to evidence brief

    Navigation example, not a model recommendation.

Visual proof

What the workflow produces

The panel below renders the five steps for the long-context-analysis demo, with every tile linking to the live surface the demo walks through. The data comes straight from the typed local data layer — no fabricated screenshots.

Workflow preview

Long-context analysis

Server-rendered from current catalogue data. Each tile links to the live surface the demo walks through.

  1. Step 1 · use case

    Long-context analysis

    Names the verified fields this workflow weights.

  2. Step 2 · shortlist

    9 candidates

    Deterministic order — no scoring.

  3. Step 3 · comparison

    4 models side by side

    Verified fields only — no derived metrics.

  4. Step 4 · brief

    6 evidence fields

    Markdown or JSON. Evidence, not recommendation.

  5. Step 5 · sources

    Citation registry

    Every value linked back to a primary source.

Demo only — the catalogue does not assert which long-context model is best for any particular workload. Inspect verified context + pricing fields, then run task-specific tests in your own environment.

Why these demos exist

  • Show the workflow end to end without forcing a visitor to discover every URL on their own.
  • Demonstrate the verified-field approach with real model records, real source citations, and real freshness states.
  • Surface explicit data gaps so the reader can see what the catalogue does not know — and where to go to confirm a value externally.
  • Make the no-ranking / no-recommendation policy obvious. The demos walk evidence; they do not point at "the right model".

After the demo, complete an exercise

Demos walk the workflow; exercises ask you to produce the artifact. Pair them in either order. When the artifact is ready, run the matching lab playbook and an evaluation prompt set before integration.