Models and credits
The LogX and Audity models, what each meters in XMC (XPLG Model Credits) per GB of the LogX Base tier, how that tier is sized from XpoLog and PortX, the evaluation limits and how to turn a model on.
What you will do: choose the models your use case needs, size the LogX Base tier they meter against, and know what the evaluation license allows.
The wiki scopes LogX "models and assistants" per edition without defining editions; the models below are dock's commercial layer, defined in the price book and shown on Usage & AI in your workspace. In the product itself LogX is one App enabled from the Marketplace; the models decide what you are licensed and metered for.
The models
| Model | Family | XMC per GB | Needs |
|---|---|---|---|
| Observability Detect | LogX | 12 | LogX Base |
| Cyber Detect | LogX | 18 | LogX Base |
| Service Intelligence | LogX | 10 | LogX Base |
| Entity Behavior (i2) | Audity | 30 | LogX Base |
| Cross-source Correlation (i2) | Audity | 45 | LogX Base |
All five meter per GB of the LogX Base tier. Turning on an Audity model adds XpoLog and LogX to the license automatically when they are missing. The GenAI features — Assistant, Insights, DevOps & Support Agent — are the "XPLG AI" family, metered per 1,000 input and output tokens, not per GB; they are not a product.
The tier they meter against
The three platform tiers nest: PortX is the origin GB/day entering the platform; XpoLog is the index share of it (50 % by default) or an explicit volume, never above PortX; LogX Base is a share of XpoLog (100 % by default) or an explicit volume, never below the 20 GB/day tier. Every model after PortX Reduce meters at most the share Reduce keeps (40 % by default). Worked example at the default state: 500 GB/day origin, XpoLog 250 GB/day, LogX Base 250 GB/day; Cyber Detect then meters 18 × 250 GB each day. See How licensing works for the tiers, packs and terms.
Evaluation and limits
- The evaluation license (30 days, created on the first open of a workspace) includes every product at evaluation limits, 100,000 XMC, one LogX node and 100 identities; nodes and identities stay usage meters, not price-book units.
- When a model's consumption exceeds the XMC pool, dock shows the balance, run rate and forecast with alerts at 50, 75, 90 and 100 %, and Buy or extend adds a pack (100,000, 500,000 or 1,000,000) or an on-demand top-up.
- In the product, System Health raises "Allowed data volume exceeded" when the licensed volume is passed; the license is one
.licper server or cluster, installed under PortX > Settings > License.
Steps: turn a model on
- Check the license on Licenses in dock: the LogX Base tier and the models it entitles.
- In the product open Apps > Marketplace and deploy LogX (build 7.8050 or later); the Audity models are licensed on the same LogX Base tier (Audity has no wiki product page yet).
- Give the engines a few normal days, then read Deep text inspection and Anomaly detection for what to expect.
- Watch Usage & AI for the daily XMC per model against the plan.
What is not published
The Deep Text Inspection percentages and the MTTR figures on xplg.com are pending confirmation from XPLG engineering and are not repeated here.
Next: Audity
Source: XPLG wiki, pages 1594131316, 1602846618, 1594133274, 1594132938 (https://xpolog.atlassian.net/wiki/spaces/XPOL/pages/1594131316); dock price book (modules/pricing/src/catalog/xplg-catalog.json).
