// MCP TOKEN COST

The Stable Baseline MCP server costs 46,634 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 Stable Baseline MCP server's 196 tool definitions consume 46,634 tokens — 23% of a 200k context window, and 22× the median MCP server (2,145 tokens). A scoped grant exposing only the tools you use cuts that roughly in proportion.

MEASURED FROM SCHEMAS tiktoken o200k_base · rank #20 of 5,253 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 23%

1M WINDOW 4.7%

Corpus context: Stable Baseline ranks #20 of 5,253 measured MCP servers by definition cost. The median is 2,145 tokens, p90 is 11,440, and the heaviest (UnClick) is 147,411 — 74% 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 46,634 tokens go.

Each row is one tool definition as a tools/list entry — name, description and input schema — counted with o200k_base. Average: 238 tokens per tool.

Tool	Category	Tokens	% of server
addWhiteboardElements	Write	4,631	9.9%
updateImprovement	Write	1,377	3.0%
updateTask	Write	1,322	2.8%
autoDesignWhiteboard	Execute	1,155	2.5%
createImprovement	Write	1,145	2.5%
designDeckInWhiteboard	Write	1,079	2.3%
designIllustrationInWhiteboard	Write	1,011	2.2%
editDocument	Write	977	2.1%
updateWhiteboardScene	Write	803	1.7%
listWhiteboardStencils	Read	792	1.7%
insertWhiteboardDiagram	Write	762	1.6%
kg_search	Read	727	1.6%
createTask	Write	709	1.5%
insertDiagramInDocument	Write	658	1.4%
startMeetingScribe	Execute	627	1.3%
designComponent	Write	623	1.3%
insertWhiteboardImage	Write	613	1.3%
updateDiagramInDocument	Write	576	1.2%
listImprovements	Read	543	1.2%
listDocuments	Read	516	1.1%
traceImage	Write	475	1.0%
getWhiteboardImage	Read	461	1.0%
getDocument	Read	457	1.0%
duplicateWhiteboardElements	Write	455	1.0%
createBrandKit	Write	448	1.0%
listAssignablePrincipals	Read	441	0.9%
searchImprovements	Read	430	0.9%
renderDiagram	Execute	399	0.9%
updateUserPreferences	Write	398	0.9%
getDeckReplyInWhiteboard	Read	386	0.8%
exportFromWhiteboard	Read	363	0.8%
listPlans	Read	360	0.8%
listArchitectureIcons	Read	355	0.8%
dataToTable	Write	349	0.7%
insertImageInDocument	Write	340	0.7%
editWhiteboardImageRegion	Write	330	0.7%
createPlanPhase	Write	323	0.7%
updatePlanPhase	Write	321	0.7%
findAndReplaceTextInDocument	Write	314	0.7%
getDiagramImage	Read	291	0.6%
listTasks	Read	291	0.6%
createDocument	Write	275	0.6%
startSignup	Other	265	0.6%
rebuildPlatformCatalogEmbeddings	Write	258	0.6%
createDocumentIngestSession	Write	250	0.5%
listTeams	Read	245	0.5%
createTaskDependency	Write	231	0.5%
inviteMember	Write	230	0.5%
searchInfographicTemplates	Read	223	0.5%
createDocumentFromUpload	Write	221	0.5%
applyTaskDependencyCascade	Write	220	0.5%
upsertResourcePermission	Write	219	0.5%
createWhiteboard	Write	218	0.5%
updatePlan	Write	209	0.4%
updateImageInDocument	Write	205	0.4%
listDocumentVersions	Read	203	0.4%
getWhiteboard	Read	200	0.4%
getProjectHierarchy	Read	198	0.4%
createPlan	Write	193	0.4%
createProject	Write	187	0.4%
getFolderHierarchy	Read	183	0.4%
listFolders	Read	183	0.4%
pollSignupStatus	Read	175	0.4%
previewKgScopeChange	Read	174	0.4%
getDiagramInDocument	Read	172	0.4%
previewSubscriptionChange	Financial	168	0.4%
reorderDocuments	Write	168	0.4%
listProjects	Read	162	0.3%
getEffectivePermission	Read	160	0.3%
getMeetingScribeStatus	Read	160	0.3%
reorderFolders	Write	160	0.3%
setResourcePermissionOverride	Write	157	0.3%
listTaskDependencies	Read	156	0.3%
updateProject	Write	156	0.3%
listResourcePermissions	Read	155	0.3%
listWorkspaces	Read	153	0.3%
addImprovementEvidence	Write	153	0.3%
deleteDiagramInDocument	Destructive	148	0.3%
getKgScopeTree	Read	147	0.3%
createOrganisation	Write	147	0.3%
listMembers	Read	145	0.3%
listInvitations	Read	144	0.3%
addImprovementActivity	Write	144	0.3%
addPlanActivity	Write	142	0.3%
createImprovementCategory	Write	141	0.3%
setDefaultBrandKit	Write	140	0.3%
updateMemberRole	Write	140	0.3%
kg_evaluate_retrieval	Read	138	0.3%
quoteCreditPackage	Financial	134	0.3%
createTeam	Write	133	0.3%
updateProfile	Write	129	0.3%
getSubscription	Read	127	0.3%
setKgProjectVisibility	Write	126	0.3%
updateOrgSettings	Write	126	0.3%
updateTaskDependency	Write	126	0.3%
removeMember	Destructive	125	0.3%
updateOrganisation	Write	124	0.3%
cancelInvitation	Write	123	0.3%
createImageUploadSession	Write	123	0.3%
setKgWorkspaceScope	Write	123	0.3%
listDiagramTypes	Read	121	0.3%
setMemberActive	Write	120	0.3%
updateTeam	Write	120	0.3%
resendInvitation	Write	118	0.3%
listCreditPurchases	Read	118	0.3%
createWorkspace	Write	117	0.3%
updateOrgFeatureFlags	Write	117	0.3%
addWorkspaceMember	Write	116	0.2%
applySubscriptionChange	Financial	116	0.2%
stopMeetingScribe	Execute	115	0.2%
createVegaDataUploadSession	Write	114	0.2%
previewSubscriptionCancellation	Financial	113	0.2%
getDocumentIngestJob	Read	112	0.2%
getImprovement	Read	108	0.2%
previewTaskDependencyCascade	Read	107	0.2%
getTeam	Read	107	0.2%
listInvoices	Read	107	0.2%
searchTools	Read	105	0.2%
updateWorkspaceMember	Write	104	0.2%
deleteResourcePermission	Destructive	103	0.2%
updateResourcePermission	Write	103	0.2%
cancelAllKgInScope	Destructive	102	0.2%
updateImprovementComment	Write	101	0.2%
previewKgRebuild	Read	100	0.2%
updatePlanComment	Write	99	0.2%
getDiagramTypeGuide	Read	98	0.2%
updateTeamWorkspaceAccess	Write	97	0.2%
getOrganisation	Read	96	0.2%
setKgDocumentScope	Write	96	0.2%
purchaseCreditPackage	Financial	95	0.2%
getWhiteboardGuide	Read	95	0.2%
setKgFolderScope	Write	95	0.2%
setPlanItemParent	Write	95	0.2%
getCurrentPlanEntitlements	Read	94	0.2%
getMember	Read	94	0.2%
grantTeamWorkspaceAccess	Write	94	0.2%
getCreditBalance	Read	93	0.2%
updateWorkspace	Write	93	0.2%
reactivateSubscription	Financial	91	0.2%
reorderPlanPhases	Write	91	0.2%
acceptTaskDependencyReview	Write	90	0.2%
getTask	Read	90	0.2%
listOrganisations	Read	89	0.2%
getOrgSettings	Read	88	0.2%
listPaymentMethods	Read	88	0.2%
reorderImprovementCategories	Write	88	0.2%
getImageInDocument	Read	87	0.2%
resetDocumentInBrain	Destructive	86	0.2%
getCurrentUser	Read	86	0.2%
getPriceForTier	Read	86	0.2%
updateImprovementCategory	Write	86	0.2%
getCustomerPortalLink	Financial	85	0.2%
cancelSubscription	Financial	83	0.2%
getProject	Read	83	0.2%
addTeamMember	Write	81	0.2%
kg_related_documents	Read	78	0.2%
applyKgScopeChange	Write	77	0.2%
updateBillingEmail	Financial	76	0.2%
getWorkspace	Read	75	0.2%
removeWorkspaceMember	Destructive	74	0.2%
listBrandKits	Read	74	0.2%
cancelKgBuildBatch	Destructive	73	0.2%
updateFolder	Write	70	0.2%
revokeTeamWorkspaceAccess	Destructive	69	0.1%
dismissTaskDependencyReview	Write	68	0.1%
removeTeamMember	Destructive	68	0.1%
getUserPreferences	Read	68	0.1%
createFolder	Write	68	0.1%
getCreditPackages	Read	67	0.1%
triggerKgRebuild	Execute	66	0.1%
deleteTeam	Destructive	64	0.1%
getPlan	Read	64	0.1%
kg_scope_status	Read	63	0.1%
deleteVegaDataFile	Destructive	62	0.1%
listWhiteboards	Read	59	0.1%
deletePlan	Destructive	57	0.1%
getPlanHierarchy	Read	56	0.1%
getPlanPhase	Read	56	0.1%
deleteImageInDocument	Destructive	55	0.1%
deletePlanPhase	Destructive	54	0.1%
kg_list_communities	Read	53	0.1%
deleteImprovementComment	Destructive	52	0.1%
deletePlanComment	Destructive	50	0.1%
kg_get_wiki_page	Read	49	0.1%
kg_get_entity	Read	48	0.1%
listImprovementCategories	Read	46	0.1%
deleteTaskDependency	Destructive	45	0.1%
deleteWhiteboard	Destructive	45	0.1%
kg_suggest_sample_questions	Read	45	0.1%
deleteImprovement	Destructive	44	0.1%
kg_backlinks	Read	44	0.1%
deleteFolder	Destructive	43	0.1%
deleteImprovementCategory	Destructive	43	0.1%
listPlanPhases	Read	42	0.1%
getCdmdLanguageGuide	Read	40	0.1%
deleteDocument	Destructive	35	0.1%

// CUT THE COST

Your agent uses a handful of these tools. It pays for all 196.

You don't need all 196 of those definitions in the window. PolicyLayer is an MCP gateway that sits in front of Stable Baseline: 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 196 tools (no gateway)	46,634 tokens	—
3 granted tools	~714 tokens	−98%
5 granted tools	~1,190 tokens	−97%
10 granted tools	~2,379 tokens	−95%

The risk dividend: 31 of these 196 tools are critical-risk (destructive or financial) and cost 2,458 tokens (5% of the definition load). Block them — the recommended starter policy — and you reclaim that context before tuning anything else.

Create a free account and register Stable Baseline — 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 STABLE BASELINE TOKEN COST →

Instant setup, no code required.

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

//FAQ

Stable Baseline token-cost questions.

How many tokens does the Stable Baseline MCP server use?+

Its 196 tool definitions total 46,634 tokens — 23% 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 Stable Baseline 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 Stable Baseline's token usage?+

Expose fewer tools. A PolicyLayer grant scopes Stable Baseline 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 714 tokens, a 98% 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 15-07-2026 from the PolicyLayer scan database over all 196 catalogued Stable Baseline tools. Counts refresh with every site build.