io.github.mcp-dir/gcp_billing-mcp
GCP Billing
Google Cloud Billing BigQuery export reporting through a user-connected service account.
alive probed 2026-10-14 · HTTP 200
9tools
2.4ktokens in Claude Opus 5.5
2.6ktokens in Claude Code
1.2 %of a 200k context window
1.3ko200k_base estimate
270 msto initialize
139 msto list tools
- Endpoint
- https://api.mcp.ai/p_gcp_billing (streamable-http)
- Host
- api.mcp.ai
- Protocol
- 2025-11-25
- Reports itself as
- mcp-ai 0.9.372
- Capabilities
- tools
- Instructions
- 511 characters
- Claude count
- Claude Opus 5.5, measured 2026-10-14. Counted again when the tool definitions change. Method.
- Repository
- https://github.com/mcp-dir/gcp_billing-mcp
- Registry
- active, updated 2026-08-21 · registry entry
- Reachable
- 11 of the last 12 probes
Most expensive tools
| Tool | Tokens (o200k) | Share of this server |
|---|---|---|
| marketplace | 702 | 52 % |
| authenticate | 126 | 9 % |
| gcp_billing_top_services | 102 | 8 % |
| gcp_billing_cost | 94 | 7 % |
| report_bug | 84 | 6 % |
History
We add a row when the result changes. A server with no change keeps its last row.
| Probed | Result | Init | Tools | Claude | o200k |
|---|---|---|---|---|---|
| 2026-10-14 09:27 | alive | 270 ms | 9 | 2.4k | 1.3k |
| 2026-10-13 11:22 | server error | ||||
| 2026-10-11 10:27 | alive | 117 ms | 9 | 2.4k | 1.3k |
| 2026-10-09 11:24 | alive | 303 ms | 9 | 2.4k | 1.3k |
| 2026-10-07 11:16 | alive | 412 ms | 9 | 2.4k | 1.3k |
| 2026-10-06 11:04 | alive | 1.0 s | 9 | 2.4k | 1.3k |
| 2026-10-05 15:04 | alive | 426 ms | 9 | 2.4k | 1.3k |
| 2026-10-05 00:39 | alive | 77 ms | 9 | 1.3k | |
| 2026-10-04 10:07 | alive | 138 ms | 9 | 1.3k | |
| 2026-10-04 08:02 | alive | 130 ms | 9 | 1.3k | |
| 2026-10-03 20:41 | alive | 144 ms | 9 | 1.3k | |
| 2026-10-03 17:58 | alive | 423 ms | 9 | 1.3k |
Badge and feed
[](https://mcpcost.ulehla.dev/s/io.github.mcp-dir/gcp_billing-mcp)
An entry for every probe that changed the result, the tool count or the token cost: Atom feed. Add it to a feed reader or a Slack RSS app to hear when the server stops answering or its tool set grows.