// MCP TOKEN COST

The AlpineDataWorks Intelligence Server MCP server costs 60,114 tokens before the first call.

Every request your agent makes carries every tool definition this server exposes — context your code, documents and conversation can't use, mostly for tools the agent never calls. You don't need them all in the window, and you don't have to pay for them.

QUICK ANSWER The AlpineDataWorks Intelligence Server MCP server's 318 tool definitions consume 60,114 tokens — 30% of a 200k context window, and 32× the median MCP server (1,905 tokens). A scoped grant exposing only the tools you use cuts that roughly in proportion.

MEASURED FROM SCHEMAS tiktoken o200k_base · rank #26 of 8,377 measured servers · refreshed every build Method →

// CONTEXT WINDOW SHARE

What that costs before your agent starts working.

Tool definitions are overhead: they occupy context on every request and compete with your code, documents and conversation history for the same window.

200K WINDOW 30%

1M WINDOW 6.0%

Corpus context: AlpineDataWorks Intelligence Server ranks #26 of 8,377 measured MCP servers by definition cost. The median is 1,905 tokens, p90 is 12,638, and the heaviest (Ainumbers Mcp Apps) is 327,608 — 164% of a 200k window on its own. New to this? See MCP token cost and context window in the glossary.

Token cost is one axis. See the risk picture across your whole stack →

// PER-TOOL BREAKDOWN

Where the 60,114 tokens go.

Each row is one tool definition as a tools/list entry — name, description and input schema — counted with o200k_base. Average: 189 tokens per tool.

Tool	Category	Tokens	% of server
adw.feedback	Write	247	0.4%
adw.adw_054	Read	220	0.4%
adw.adw_055	Read	215	0.4%
adw.adw_563	Execute	213	0.4%
adw.adw_053	Read	213	0.4%
adw.adw_102	Read	213	0.4%
adw.adw_046	Read	212	0.4%
adw.adw_037	Read	211	0.4%
adw.adw_434	Read	210	0.3%
adw.adw_023	Read	209	0.3%
adw.adw_105	Read	209	0.3%
adw.adw_578	Read	209	0.3%
adw.adw_049	Read	207	0.3%
adw.adw_549	Read	207	0.3%
adw.adw_213	Read	206	0.3%
adw.adw_252	Read	206	0.3%
adw.adw_566	Read	206	0.3%
adw.adw_568	Read	206	0.3%
adw.adw_570	Read	206	0.3%
adw.adw_589	Read	206	0.3%
adw.adw_017	Read	205	0.3%
adw.adw_540	Read	205	0.3%
adw.adw_593	Read	205	0.3%
adw.adw_076	Read	204	0.3%
adw.adw_106	Read	204	0.3%
adw.adw_564	Read	204	0.3%
adw.adw_008	Read	203	0.3%
adw.adw_015	Read	203	0.3%
adw.adw_047	Read	203	0.3%
adw.adw_524	Execute	202	0.3%
adw.adw_003	Read	202	0.3%
adw.adw_042	Read	202	0.3%
adw.adw_071	Read	202	0.3%
adw.adw_072	Read	202	0.3%
adw.adw_104	Read	202	0.3%
adw.adw_418	Read	202	0.3%
adw.adw_560	Read	202	0.3%
adw.adw_579	Read	202	0.3%
adw.adw_619	Read	202	0.3%
adw.adw_073	Read	201	0.3%
adw.adw_101	Read	201	0.3%
adw.adw_212	Read	201	0.3%
adw.adw_335	Read	201	0.3%
adw.adw_007	Read	200	0.3%
adw.adw_411	Read	200	0.3%
adw.adw_426	Read	200	0.3%
adw.adw_428	Read	200	0.3%
adw.adw_430	Read	200	0.3%
adw.adw_596	Read	200	0.3%
adw.adw_557	Execute	199	0.3%
adw.adw_026	Read	199	0.3%
adw.adw_044	Read	199	0.3%
adw.adw_045	Read	199	0.3%
adw.adw_552	Read	199	0.3%
adw.adw_592	Read	199	0.3%
adw.adw_601	Read	199	0.3%
adw.adw_609	Read	199	0.3%
adw.adw_001	Read	198	0.3%
adw.adw_052	Read	198	0.3%
adw.adw_103	Read	198	0.3%
adw.adw_347	Read	198	0.3%
adw.adw_349	Read	198	0.3%
adw.adw_414	Read	198	0.3%
adw.adw_569	Read	198	0.3%
adw.adw_610	Read	198	0.3%
adw.adw_505	Execute	197	0.3%
adw.adw_034	Read	197	0.3%
adw.adw_038	Read	197	0.3%
adw.adw_056	Read	197	0.3%
adw.adw_132	Read	197	0.3%
adw.adw_250	Read	197	0.3%
adw.adw_251	Read	197	0.3%
adw.adw_355	Read	197	0.3%
adw.adw_422	Read	197	0.3%
adw.adw_518	Read	197	0.3%
adw.adw_539	Read	197	0.3%
adw.adw_575	Read	197	0.3%
adw.adw_576	Read	197	0.3%
adw.adw_588	Read	197	0.3%
adw.adw_594	Read	197	0.3%
adw.adw_608	Read	197	0.3%
adw.adw_p18	Read	197	0.3%
adw.adw_208	Read	196	0.3%
adw.adw_253	Read	196	0.3%
adw.adw_311	Read	196	0.3%
adw.adw_357	Read	196	0.3%
adw.adw_419	Read	196	0.3%
adw.adw_525	Read	196	0.3%
adw.adw_555	Read	196	0.3%
adw.adw_006	Read	195	0.3%
adw.adw_028	Read	195	0.3%
adw.adw_050	Read	195	0.3%
adw.adw_122	Read	195	0.3%
adw.adw_217	Read	195	0.3%
adw.adw_559	Read	195	0.3%
adw.adw_562	Read	195	0.3%
adw.adw_581	Read	195	0.3%
adw.adw_602	Read	195	0.3%
adw.adw_603	Read	195	0.3%
adw.adw_615	Read	195	0.3%
adw.adw_618	Read	195	0.3%
adw.adw_p06	Read	195	0.3%
adw.adw_031	Read	194	0.3%
adw.adw_040	Read	194	0.3%
adw.adw_112	Read	194	0.3%
adw.adw_254	Read	194	0.3%
adw.adw_412	Read	194	0.3%
adw.adw_413	Read	194	0.3%
adw.adw_435	Read	194	0.3%
adw.adw_512	Read	194	0.3%
adw.adw_572	Read	194	0.3%
adw.adw_590	Read	194	0.3%
adw.adw_591	Read	194	0.3%
adw.adw_595	Read	194	0.3%
adw.adw_024	Read	193	0.3%
adw.adw_067	Read	193	0.3%
adw.adw_207	Read	193	0.3%
adw.adw_313	Read	193	0.3%
adw.adw_354	Read	193	0.3%
adw.adw_415	Read	193	0.3%
adw.adw_432	Read	193	0.3%
adw.adw_553	Read	193	0.3%
adw.adw_574	Read	193	0.3%
adw.adw_583	Read	193	0.3%
adw.adw_607	Read	193	0.3%
adw.adw_611	Read	193	0.3%
adw.adw_125	Read	192	0.3%
adw.adw_223	Read	192	0.3%
adw.adw_356	Read	192	0.3%
adw.adw_382	Read	192	0.3%
adw.adw_416	Read	192	0.3%
adw.adw_533	Read	192	0.3%
adw.adw_573	Read	192	0.3%
adw.adw_123	Read	191	0.3%
adw.adw_133	Read	191	0.3%
adw.adw_206	Read	191	0.3%
adw.adw_339	Read	191	0.3%
adw.adw_383	Read	191	0.3%
adw.adw_395	Read	191	0.3%
adw.adw_417	Read	191	0.3%
adw.adw_439	Read	191	0.3%
adw.adw_526	Read	191	0.3%
adw.adw_537	Read	191	0.3%
adw.adw_545	Read	191	0.3%
adw.adw_604	Read	191	0.3%
adw.adw_612	Read	191	0.3%
adw.adw_550	Execute	190	0.3%
adw.adw_010	Read	190	0.3%
adw.adw_033	Read	190	0.3%
adw.adw_129	Read	190	0.3%
adw.adw_130	Read	190	0.3%
adw.adw_131	Read	190	0.3%
adw.adw_202	Read	190	0.3%
adw.adw_384	Read	190	0.3%
adw.adw_392	Read	190	0.3%
adw.adw_397	Read	190	0.3%
adw.adw_408	Read	190	0.3%
adw.adw_424	Read	190	0.3%
adw.adw_528	Read	190	0.3%
adw.adw_605	Read	190	0.3%
adw.adw_616	Read	190	0.3%
adw.adw_233	Read	189	0.3%
adw.adw_377	Read	189	0.3%
adw.adw_409	Read	189	0.3%
adw.adw_508	Read	189	0.3%
adw.adw_520	Read	189	0.3%
adw.adw_543	Read	189	0.3%
adw.adw_554	Read	189	0.3%
adw.adw_558	Read	189	0.3%
adw.adw_565	Read	189	0.3%
adw.adw_577	Read	189	0.3%
adw.adw_597	Read	189	0.3%
adw.adw_598	Read	189	0.3%
adw.adw_600	Read	189	0.3%
adw.adw_606	Read	189	0.3%
adw.adw_p03	Read	189	0.3%
adw.adw_002	Read	188	0.3%
adw.adw_030	Read	188	0.3%
adw.adw_210	Read	188	0.3%
adw.adw_219	Read	188	0.3%
adw.adw_314	Read	188	0.3%
adw.adw_400	Read	188	0.3%
adw.adw_431	Read	188	0.3%
adw.adw_506	Read	188	0.3%
adw.adw_534	Read	188	0.3%
adw.adw_544	Read	188	0.3%
adw.adw_548	Read	188	0.3%
adw.adw_580	Read	188	0.3%
adw.adw_587	Read	188	0.3%
adw.adw_p08	Read	188	0.3%
adw.adw_009	Read	187	0.3%
adw.adw_018	Read	187	0.3%
adw.adw_115	Read	187	0.3%
adw.adw_124	Read	187	0.3%
adw.adw_215	Read	187	0.3%
adw.adw_258	Read	187	0.3%
adw.adw_348	Read	187	0.3%
adw.adw_385	Read	187	0.3%
adw.adw_402	Read	187	0.3%
adw.adw_404	Read	187	0.3%
adw.adw_436	Read	187	0.3%
adw.adw_536	Read	187	0.3%
adw.adw_551	Read	187	0.3%
adw.adw_556	Read	187	0.3%
adw.adw_020	Read	186	0.3%
adw.adw_110	Read	186	0.3%
adw.adw_232	Read	186	0.3%
adw.adw_255	Read	186	0.3%
adw.adw_261	Read	186	0.3%
adw.adw_310	Read	186	0.3%
adw.adw_312	Read	186	0.3%
adw.adw_386	Read	186	0.3%
adw.adw_390	Read	186	0.3%
adw.adw_502	Read	186	0.3%
adw.adw_511	Read	186	0.3%
adw.adw_514	Read	186	0.3%
adw.adw_527	Read	186	0.3%
adw.adw_535	Read	186	0.3%
adw.adw_546	Read	186	0.3%
adw.adw_p04	Read	186	0.3%
adw.adw_062	Read	185	0.3%
adw.adw_075	Read	185	0.3%
adw.adw_116	Read	185	0.3%
adw.adw_128	Read	185	0.3%
adw.adw_209	Read	185	0.3%
adw.adw_220	Read	185	0.3%
adw.adw_387	Read	185	0.3%
adw.adw_500	Read	185	0.3%
adw.adw_503	Read	185	0.3%
adw.adw_571	Read	185	0.3%
adw.adw_613	Execute	184	0.3%
adw.adw_032	Read	184	0.3%
adw.adw_238	Read	184	0.3%
adw.adw_393	Read	184	0.3%
adw.adw_513	Read	184	0.3%
adw.adw_541	Read	184	0.3%
adw.adw_p15	Read	184	0.3%
adw.adw_p24	Read	184	0.3%
adw.adw_504	Execute	183	0.3%
adw.adw_p14	Execute	183	0.3%
adw.adw_005	Read	183	0.3%
adw.adw_029	Read	183	0.3%
adw.adw_121	Read	183	0.3%
adw.adw_401	Read	183	0.3%
adw.adw_405	Read	183	0.3%
adw.adw_501	Read	183	0.3%
adw.adw_517	Read	183	0.3%
adw.adw_519	Read	183	0.3%
adw.adw_532	Read	183	0.3%
adw.adw_561	Read	183	0.3%
adw.adw_582	Read	183	0.3%
adw.adw_614	Read	183	0.3%
adw.adw_617	Read	183	0.3%
adw.adw_p17	Read	183	0.3%
adw.adw_262	Read	182	0.3%
adw.adw_263	Read	182	0.3%
adw.adw_388	Read	182	0.3%
adw.adw_507	Read	182	0.3%
adw.adw_509	Read	182	0.3%
adw.adw_515	Read	182	0.3%
adw.adw_516	Read	182	0.3%
adw.adw_529	Read	182	0.3%
adw.adw_547	Read	182	0.3%
adw.adw_584	Read	182	0.3%
adw.adw_p09	Read	182	0.3%
adw.adw_p23	Read	182	0.3%
adw.adw_204	Read	181	0.3%
adw.adw_259	Read	181	0.3%
adw.adw_264	Read	181	0.3%
adw.adw_394	Read	181	0.3%
adw.adw_523	Read	181	0.3%
adw.adw_542	Read	181	0.3%
adw.adw_599	Read	181	0.3%
adw.adw_p10	Read	181	0.3%
adw.adw_p12	Read	181	0.3%
adw.adw_p16	Read	181	0.3%
adw.adw_004	Read	180	0.3%
adw.adw_027	Read	180	0.3%
adw.adw_201	Read	180	0.3%
adw.adw_205	Read	180	0.3%
adw.adw_256	Read	180	0.3%
adw.adw_260	Read	180	0.3%
adw.adw_510	Read	180	0.3%
adw.adw_522	Read	180	0.3%
adw.adw_p07	Read	180	0.3%
adw.adw_p19	Read	180	0.3%
adw.adw_p22	Read	180	0.3%
adw.adw_234	Read	179	0.3%
adw.adw_391	Read	179	0.3%
adw.adw_p05	Read	179	0.3%
adw.adw_p13	Read	179	0.3%
adw.adw_016	Read	178	0.3%
adw.adw_021	Read	178	0.3%
adw.adw_531	Read	178	0.3%
adw.adw_p02	Read	178	0.3%
adw.adw_127	Read	177	0.3%
adw.adw_521	Read	177	0.3%
adw.adw_538	Read	177	0.3%
adw.adw_p01	Read	177	0.3%
adw.county_cancer	Read	176	0.3%
adw.adw_p21	Read	175	0.3%
adw.adw_041	Read	174	0.3%
adw.adw_117	Read	173	0.3%
adw.adw_120	Read	173	0.3%
adw.adw_126	Read	173	0.3%
adw.adw_257	Read	173	0.3%
adw.adw_530	Read	173	0.3%
adw.adw_p11	Read	173	0.3%
adw.air_quality_risk	Read	172	0.3%
adw.adw_019	Read	170	0.3%
adw.county_sdoh	Read	157	0.3%
adw.county_mortality	Read	150	0.2%
adw.county_water	Read	143	0.2%
adw.obesity_risk	Read	140	0.2%
adw.adw_901	Read	112	0.2%
adw.health_risk	Read	95	0.2%
adw.sample	Read	87	0.1%
adw.catalog	Read	60	0.1%

// CUT THE COST

Your agent uses a handful of these tools. It pays for all 318.

You don't need all 318 of those definitions in the window. PolicyLayer is an MCP gateway that sits in front of AlpineDataWorks Intelligence Server: only the tools you grant are exposed to the agent, the rest never load. A smaller window means a sharper agent — less noise when it picks a tool — and every request costs less:

Grant scope	Definition cost	Reduction
All 318 tools (no gateway)	60,114 tokens	—
3 granted tools	~567 tokens	−99%
5 granted tools	~945 tokens	−98%
10 granted tools	~1,890 tokens	−97%

Create a free account and register AlpineDataWorks Intelligence Server — nothing to install.
Grant only the tools you use — ungranted definitions never enter the context window.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

CUT ALPINEDATAWORKS INTELLIGENCE TOKEN COST →

Instant setup, no code required.

Model your own stack in the token-cost calculator, or see the AlpineDataWorks Intelligence Server policy for what a sensible grant looks like.

//FAQ

AlpineDataWorks Intelligence Server token-cost questions.

How many tokens does the AlpineDataWorks Intelligence Server MCP server use?+

Its 318 tool definitions total 60,114 tokens — 30% of a 200k context window — measured with tiktoken o200k_base over the serialised tools/list payload. Exact counts vary slightly by client and model.

Why does AlpineDataWorks Intelligence Server consume tokens before I send a message?+

MCP clients load every connected server's tool definitions — name, description, and input schema — into the model's context so it knows what it can call. That payload is charged against your context window on every request, whether or not a tool is used.

How do I reduce AlpineDataWorks Intelligence Server's token usage?+

Expose fewer tools. A PolicyLayer grant scopes AlpineDataWorks Intelligence Server to only the tools you allow — ungranted definitions are filtered out of the tool list, so they never enter the context window. A grant of 3 typical tools costs roughly 567 tokens, a 99% reduction.

Does deferred tool loading fix this?+

Partially, in some clients. Claude Code defers MCP tool schemas behind a tool-search step by default, and VS Code has experimental grouping — but you still pay tokens per search and reload, and Cursor, Windsurf and Gemini CLI load definitions upfront. Reducing the exposed tool set cuts the cost in every client.

// METHOD & REVIEW

How these numbers were measured.

01

Serialisation

Each tool is serialised as a tools/list entry — name, description, input schema — from the schemas in the PolicyLayer scan database. Clients differ slightly in framing, so treat counts as close estimates.

02

Tokeniser

tiktoken o200k_base (GPT-4o/o-series). Anthropic's current tokeniser isn't published, so Claude's exact counts will differ; for English text and JSON schemas the totals are close enough to treat these as estimates.

03

Deferred loading

Some clients now defer schema loading (Claude Code's tool search; VS Code experimental grouping). You still pay per search and reload — and Cursor, Windsurf and Gemini CLI load everything upfront.

Computed 06-09-2026 from the PolicyLayer scan database over all 318 catalogued AlpineDataWorks Intelligence Server tools. Counts refresh with every site build.