Can My Agent Uselookup · compare · cite

Aug 28, 2026 · Catalog foundation is up

Product feature · operations

Usage meteringCommon product term

Inspect documented product usage such as requests, tokens, or cost.

Terminology basis: Common product term. Anthropic — Manage Claude Code costs effectively

Usage metering: 4 supported, 2 partial, 0 unsupported, 25 unreviewed across 31 cataloged products.

Markdown · JSON

Explore this familyMore in Pricing, usage, and limits3 capabilities

Current evidence by product

Can my agent use Usage metering?

Read across for the answer. 6 of 31 current product columns have reviewed evidence; unreviewed does not mean unsupported.

Known issues (0)Report an issue
  • Supported4
  • Partial2
  • Unsupported0
  • Unknown25
  • Not applicable0

Web

9 products

Desktop

13 products

CLI

9 products

Unknown means no public evidence has been reviewed for that product and capability. It does not mean unsupported.

How statuses are assigned

Definition and scope

What this capability means

This row asks whether operators can see consumption close enough to a run to diagnose and control cost. Useful meters separate input, output, reasoning, and cached tokens; tool and connector calls; hosted sandbox or GPU time; storage; media processing; and parent versus sub-agent usage.

Evidence should record update delay, estimated versus final values, price mapping, currencies, user/project/organization attribution, export API, budgets and alerts, model routing, and treatment of retries and failed runs. A monthly invoice total is only partial operational visibility.

Traceable compatibility

Assertion ledger

Documentation evidence only. No runtime conformance test is implied.

ChatGPTweb · current
Supported
Target
2026-08-28 ChatGPT usage documentation observation · hosted-observation
Environment
hosted-default
Observed
2026-08-28
  • runtimethe account usage dashboard shows current limits and remaining capacity; the published rate card separates input, cached input, and output credits by model
  • planChatGPT Work and Codex share usage; included limits, flexible credits, and overage options vary by plan
Evidence
Codex CLIcli · current
Supported
Target
2026-08-28 Codex CLI usage documentation observation · hosted-observation
Environment
local-default
Observed
2026-08-28
  • runtime/status reports remaining plan limits during an active session, and the account dashboard provides the broader current allowance
  • authChatGPT-plan authentication uses shared plan limits and credits; API-key authentication is billed separately at API token rates
  • policycontext, reasoning, tools, retrieval, caching, execution mode, and model choice affect consumption, so prompt length alone does not predict usage
Evidence
Claude CLIcli · current
Supported
Target
2026-08-28 Claude Code documentation observation · hosted-observation
Environment
local-default
Observed
2026-08-28
  • runtime/usage shows session tokens by model, cache reads and writes, duration, code changes, and locally estimated cost; /clear resets session totals
  • planPro, Max, Team, and Enterprise add plan-usage bars and approximate local attribution across skills, subagents, plugins, MCP servers, and scheduled tasks
  • policysession cost is an estimate unless organization pricing is configured; authoritative billing remains in the applicable console or provider
Evidence
Cursordesktop · current
Supported
Target
2026-08-28 Cursor documentation observation · hosted-observation
Environment
local-default
Observed
2026-08-28
  • runtimeSpending dashboard shows real-time use, remaining allowance, on-demand charges, reset date, and request-level cost and pool details
  • planCursor Models and third-party Other Models draw from separate monthly pools; Auto requests are charged at the routed model's list price
  • policyportable telemetry export and separate tool, compute, storage, and subagent consumption are not established by the reviewed page
Evidence
Copilot CLIcli · current
Partial
Target
current GitHub Copilot CLI command reference · dated-documentation
Environment
local-default
Observed
2026-08-28
  • runtime/usage is a documented informational command and --max-ai-credits sets a soft per-response ceiling; the reviewed page does not enumerate the meter's dimensions, delay, or billing-final status
Evidence
Copilot Chatdesktop · current
Partial
Target
current VS Code agent-session documentation · dated-documentation
Environment
local-default
Observed
2026-08-28
  • runtimethe session context control exposes total AI credits consumed, but the reviewed page does not establish request-level cost dimensions, export, or billing-final values
Evidence
  1. 1. Evidence checked 2026-08-28: Claude Code's `/usage` screen reports current-session tokens by model, cache reads and writes, duration, code changes, and an estimated cost. Subscription users also receive plan-usage bars and attribution across skills, subagents, plugins, MCP servers, and scheduled tasks, with documented version and device-scope boundaries.
  2. 2. Evidence checked 2026-08-28: Cursor's Spending dashboard reports real-time consumption for its separate monthly model pools, remaining allowance, on-demand charges, and request-level cost and pool details. The reviewed page does not establish a portable telemetry export or tool and subagent cost breakdown.
  3. 3. Evidence checked 2026-08-28: ChatGPT's usage dashboard reports current limits and remaining capacity, while Codex CLI also exposes remaining limits through `/status`. OpenAI publishes a credit rate card separating input, cached input, and output consumption by model, but task-level consumption varies with context, reasoning, tools, retrieval, and caching.
  4. 4. Evidence checked 2026-08-28: GitHub Copilot CLI exposes `/usage` as a built-in informational command for current usage, and its model picker and credit-limit controls document session-level consumption boundaries.
  5. 5. Evidence checked 2026-08-28: VS Code's agent-session context control reports total AI credits consumed by the current session alongside context-token usage.