// MCP TOKEN COST

The Crypto Intelligence MCP server costs 8,828 tokens before the first call.

Connect Crypto Intelligence and its 104 tool definitions are loaded into the model's context on every request — 4.4% of a 200k window spent before your agent does anything.

QUICK ANSWER The Crypto Intelligence MCP server's tool definitions consume 8,828 tokens — 4.6× the median MCP server (1,905 tokens). A scoped grant exposing only the tools you use cuts that roughly in proportion.

MEASURED FROM SCHEMAS 104 tools · 8,828 tokens · 4.4% of 200k · 0.9% of 1M Method →

// CONTEXT WINDOW SHARE

What that buys 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 4.4%

1M WINDOW 0.9%

Corpus context: Crypto Intelligence ranks #205 of 3,213 measured MCP servers by definition cost. The median is 1,905 tokens, p90 is 7,952, and the heaviest (Fusionauth) is 183,337 — 92% of a 200k window on its own.

// PER-TOOL BREAKDOWN

Where the 8,828 tokens go.

Each row is one tool definition as a tools/list entry — name, description and input schema — counted with o200k_base. Average: 85 tokens per tool.

Tool	Category	Tokens	% of server
get_funding_history	Read	176	2.0%
run_strategy_test	Execute	149	1.7%
get_intel_history	Read	141	1.6%
get_subnet_leaderboard	Read	141	1.6%
get_anomaly_history	Read	138	1.6%
get_signals	Read	134	1.5%
get_oi_history	Read	130	1.5%
ask_assistant	Read	129	1.5%
get_intel_snapshot	Read	129	1.5%
get_regime_history	Read	118	1.3%
get_defi_yields	Read	115	1.3%
get_trend	Read	111	1.3%
get_tech_indicators	Read	110	1.2%
get_mean_rev	Read	105	1.2%
get_deriv_analytics	Read	104	1.2%
get_tao_burn_history	Read	104	1.2%
get_analytics_bundle	Read	103	1.2%
get_volatility	Read	102	1.2%
get_coin_research	Read	101	1.1%
get_network_stats	Read	100	1.1%
get_news_sentiment	Read	97	1.1%
get_signal_feed	Read	95	1.1%
get_decision_pack	Read	94	1.1%
get_factor_analytics	Read	94	1.1%
get_correlations	Read	92	1.0%
get_portfolio_risk	Read	92	1.0%
get_proven_record	Read	89	1.0%
get_crypto_news	Read	88	1.0%
get_intel_bundle	Read	87	1.0%
get_intel_coin	Read	87	1.0%
get_subnets	Read	87	1.0%
get_bittensor_tao	Read	86	1.0%
get_cointegration	Read	86	1.0%
get_prices	Read	86	1.0%
get_vol_regime	Read	86	1.0%
get_beta_to_btc	Read	85	1.0%
get_orderbook	Read	85	1.0%
get_hurst	Read	84	1.0%
get_kaspa_block_prop	Read	84	1.0%
get_anomaly_coin	Read	83	0.9%
get_blocks	Read	83	0.9%
get_btc_fee_market	Read	83	0.9%
get_garch	Read	83	0.9%
get_macro_corr	Read	83	0.9%
get_bittensor_epoch	Read	82	0.9%
get_btc_onchain	Read	82	0.9%
get_chart_vision	Read	82	0.9%
get_kaspa_network	Read	82	0.9%
get_macro_data	Read	82	0.9%
get_model_perf	Read	82	0.9%
get_ohlcv	Read	82	0.9%
get_onchain_pulse	Read	82	0.9%
get_gas_prices	Read	81	0.9%
get_kaspa_dag_pulse	Read	81	0.9%
get_ml_signal	Read	81	0.9%
get_coin_days	Read	79	0.9%
get_rolling_corr	Read	79	0.9%
get_conviction	Read	78	0.9%
run_backtest	Execute	77	0.9%
get_bittensor	Read	77	0.9%
get_fear_greed	Read	77	0.9%
get_mempool_fees	Read	77	0.9%
get_signal_accuracy	Read	77	0.9%
get_signal_feed_coin	Read	77	0.9%
get_utxo_set	Read	77	0.9%
scan_arbitrage	Read	77	0.9%
get_bittensor_validators	Read	76	0.9%
get_liquidations	Read	76	0.9%
get_ml_features	Read	76	0.9%
get_token_unlocks	Read	76	0.9%
get_exchange_flows	Read	75	0.8%
get_mempool	Read	75	0.8%
get_open_interest	Read	75	0.8%
get_anomalies	Read	74	0.8%
get_eth_beacon	Read	74	0.8%
get_stablecoins	Read	74	0.8%
get_whale_flow	Read	74	0.8%
get_consensus	Read	72	0.8%
get_long_short	Read	72	0.8%
get_subtensor	Read	72	0.8%
get_track_record	Read	72	0.8%
get_eco_watch	Read	71	0.8%
get_funding_rates	Read	71	0.8%
get_market_regime	Read	71	0.8%
get_options	Read	70	0.8%
get_strategies	Read	70	0.8%
get_chain_activity	Read	69	0.8%
get_daily_report	Read	69	0.8%
get_narratives	Read	68	0.8%
get_onchain_fg	Read	68	0.8%
get_market_score	Read	67	0.8%
get_weekly_report	Read	67	0.8%
get_dispersion	Read	66	0.7%
get_kaspa_mining_econ	Read	65	0.7%
get_decisions	Read	64	0.7%
get_research_replica	Read	60	0.7%
get_btc_rbf_pulse	Read	54	0.6%
get_onchain_bundle	Read	54	0.6%
get_btc_whales	Read	53	0.6%
get_btc_ultimate	Read	52	0.6%
get_quant_suite	Read	50	0.6%
get_rel_strength	Read	50	0.6%
get_sector_rotation	Read	49	0.6%
get_events_feed	Read	47	0.5%

// SCOPED GRANT

Most agents use a handful of these tools. They pay for all 104.

A PolicyLayer grant exposes only the tools you allow — ungranted definitions are filtered out of the tool list, so they never enter the context window. Estimates below assume typical-weight tools (85 tokens each).

Grant scope	Definition cost	Reduction
All 104 tools (no gateway)	8,828 tokens	—
3 granted tools	~255 tokens	−97%
5 granted tools	~424 tokens	−95%
10 granted tools	~849 tokens	−90%

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

//FAQ

Crypto Intelligence token-cost questions.

How many tokens does the Crypto Intelligence MCP server use?+

Its 104 tool definitions total 8,828 tokens — 4.4% 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 Crypto Intelligence 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 Crypto Intelligence's token usage?+

Expose fewer tools. A PolicyLayer grant scopes Crypto Intelligence 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 255 tokens, a 97% 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 07-06-2026 from the PolicyLayer scan database over all 104 catalogued Crypto Intelligence tools. Counts refresh with every site build.