Cite as: Real Problem AI problem “Why can't a team see which engineer (and which model) is actually burning the AI budget?”. Opportunity score 6.8 out of 10 (severity 6, AI feasibility 8, market signal 6, competition gap 7). Category AI / Agents. Trend LLM. Source signal: anthropics/claude-code GitHub issue #29123, 26 Feb 2026, closed as duplicate.. Canonical URL: https://www.realproblem.ai/archive/why-cant-a-team-see-which-engineer-is-burning-the-model-budget.
Why can't a team see which engineer (and which model) is actually burning the AI budget?
There's no per-user, per-model cost breakdown in Claude Code, so engineering leads can't tell whether it's Opus-heavy refactors or Sonnet-heavy chat loops driving the bill.
Who has it: Engineering managers rolling out Claude Code across a team.
Evidence
“I would like to see at a per user level how much cost I have incurred for each model. Would help me decide how to best plan my tasks”
Quoted word for word from the public post linked below. Nobody submitted it to Real Problem AI.
anthropics/claude-code GitHub issue #29123, 26 Feb 2026, closed as duplicate.Why it is archived
Trimmed to 100-cap (lowest opportunity_score)
Scoring breakdown
Existing players
- Org-level console analytics · Aggregates across the whole org, not broken down per user/model.
- Manual spreadsheet tracking · Teams self-report usage, unreliable and stale.
What they are missing
A per-user, per-model cost breakdown surfaced to both the individual (to self-regulate) and the team lead (to plan task allocation and model choice).
Stack hint
#ATB7 · Canonical URL: https://www.realproblem.ai/archive/why-cant-a-team-see-which-engineer-is-burning-the-model-budget