TL;DR Revenium 2.11.0 adds a new Claude Cowork Dashboard with per-user value reporting and zero-setup OTLP instrumentation, improves AI Coding Dashboard attribution with a GitHub integration, and introduces scoped API key permissions. You also get environment filtering across Logs and Traces, a unified Go metering SDK, clickable ROI drilldowns, deeper MCP platform upgrades, and a cleaner Pricing sidebar.
Version 2.11.0 is live! This release brings new Claude Cowork analytics and stronger attribution, plus scoped API keys and environment filtering for production teams.
Here’s what’s new.
1) Claude Cowork Dashboard (with per-user value + one-line OTLP setup)
If your company is using Claude Cowork, you can now see adoption, usage, and cost clearly at the individual level.
What’s included:
- Per-user value reporting so you can understand where Cowork is creating outcomes
- Cost by model to track spend shifts as teams adopt new models
- Token + cache utilization breakdowns to diagnose efficiency (and waste)
- One-line OTLP configuration from the Claude Desktop admin UI that instruments your entire company’s Cowork usage with zero per-machine setup
The result is fast, org-wide visibility without the usual rollout overhead.
2) GitHub integration for AI Coding Dashboard team attribution
We added GitHub team attribution so the AI Coding Dashboard can map Claude Code usage to teams based on PR authorship.
This makes it possible to:
- Attribute Claude Code usage to the teams responsible for the work
- Compare usage against meaningful outcomes (shipped code), instead of treating coding assistants as an undifferentiated “engineering” line item
- Identify which team members are utilizing Claude efficiently, and where to invest in enablement
3) Scoped API key permissions (read-only, metering-only, or full access)
You can now create API keys with explicit scopes depending on what an integration needs.
New key types include:
- Read-only
- Metering-only
- Full access
Legacy keys continue to work unchanged, so you can adopt scopes incrementally.
4) Environment filtering across Logs and Traces
Revenium now supports environment filtering across Logs and Traces so you can segment usage by dev, staging, prod, and more.
This helps teams:
- Separate test and experimentation spend from production spend
- Catch staging regressions before they become production incidents
- Answer finance questions without manual tagging cleanups
5) Unified Go metering SDK (first-class multi-provider support)
We shipped a unified Go metering SDK with first-class support for:
- Anthropic (streaming + vision)
- Google (chat, embeddings, image, video)
- fal.ai (Run, Subscribe, Stream)
- LiteLLM (including SSE streaming)
- Responses API
This makes it easier to standardize metering across providers while keeping implementation clean and consistent.
6) Clickable ROI cards (drill from ROI → job details)
ROI reporting is now interactive. You can click any box on the ROI page to drill into detailed job filtering.
That means less “looks good in aggregate” and more “show me the exact jobs behind this number.”
7) Platform upgrades (MCP Server + Tool Registry + expanded analytics)
We updated the MCP Server to FastMCP 3.x and shipped a set of platform-layer improvements:
- Tool Registry management with pricing tiers
- Jobs and Outcomes management
- Expanded analytics wiring for cost-by-user and tool costs
Together, these upgrades strengthen the foundation for deeper control and attribution as teams scale usage across more tools, users, and workflows.
As always, thanks to the engineering team for another strong push forward.
Prefer to self-serve? Watch the Revenium platform walkthrough playlist on YouTube.
Or contact us and we will share a short demo.
Not using Revenium yet? Get started for free!
