Model consumption analysis
To see how different AI models are used across your organization, open the Model consumption analysis report. It shows each user's model preferences, for example Opus, Sonnet, and Haiku, alongside a detailed list of token volumes and costs. Use it to understand model adoption and to audit AI spend.
The data comes from OpenLM AI Proxy, which you deploy in your network and connect to OpenLM Platform.
Visualizations
The report has the following visualizations:
- Token-based model mix by user bar chart: A horizontal stacked bar chart of the tokens each user consumed, segmented by the models they used. It helps identify power users, their model-tier preferences, and adoption trends. Values displayed:
- Y-axis: username
- X-axis: tokens consumed (total tokens)
- Color: model name
- Model consumption details table: A granular table of user-and-model consumption records, with licensing context, recent activity, the input and output token split, and the resulting cost. Suited to FinOps auditing and data exports. Values displayed:
- Model name
- User
- Type of license allocated (derived from feature name)
- Last active date
- Has usage (Yes or No, indicating whether compute was consumed)
- Total input tokens
- Total output tokens
- Total token consumption
- Total cost consumption
The documentation team might draft this page using a large language model. A technical writer reviews every page and verifies it against the product before publishing it. A human makes the final decision.