Optional request parameters

The sync and async clients accept the same filters. The additions below are keyword-only and default to None. Existing argument order, positional calls, defaults, and return types are preserved. Unset additions are omitted before request validation, so existing endpoint defaults still apply: for example, market.search_by_cik("320193") still requests limit=50.

Pass False or 0 explicitly when needed; neither is treated as unset. Limits and pages are sent to FMP as requested. The SDK does not automatically paginate, deduplicate overlapping pages, widen date ranges, or repair calendar records.

from datetime import date

from fmp_data import FMPDataClient

with FMPDataClient.from_env() as client:
    holidays = client.market.get_holidays_by_exchange(
        "NYSE", from_date=date(2026, 1, 1), to_date=date(2026, 12, 31)
    )
    companies = client.market.search_by_cik("320193", limit=2)
    prices = client.company.get_intraday_prices(
        "AAPL",
        "5min",
        from_date=date(2026, 9, 10),
        to_date=date(2026, 9, 11),
        extended=False,
    )
    valuation = client.fundamental.get_custom_discounted_cash_flow(
        "MSFT", risk_free_rate=3.5, revenue_growth_pct=5.0
    )

Use await client.<group>.<method>(...) with AsyncFMPDataClient and the same arguments. Dates are datetime.date; timestamps are Unix seconds. Country and SIC codes are strings. FMP documents invalid as a string, so use "false" or "true" rather than a Python boolean. Country values follow FMP's vocabulary; the audit verified US and UK, while GB returned no rows.

Added controls

Provider documentation checked on 2026-09-13. Each row links to its source.

Client method	New Python arguments	FMP query keys
market.get_holidays_by_exchange	from_date, to_date	from, to
market.get_company_screener	avg_volume_more_than, avg_volume_less_than	avgVolumeMoreThan, avgVolumeLowerThan
market.get_market_hours	timestamp	timestamp
market.get_all_exchange_market_hours	timestamp	timestamp
market.get_all_shares_float	page, limit	page, limit
market.get_available_exchanges	extended	extended
market.search_by_cik	limit	limit
company.get_employee_count	limit	limit
company.get_historical_market_cap	from_date, to_date, limit	from, to, limit
company.get_symbol_changes	invalid, limit	invalid, limit
company.get_intraday_prices	extended	extended
company.get_earnings	include_report_times	includeReportTimes
economics.get_economic_indicators	start_date, end_date	from, to
economics.get_economic_calendar	country	country
fundamental.get_custom_discounted_cash_flow	18 custom DCF assumptions (below)	Corresponding camelCase keys (below)
fundamental.get_custom_levered_dcf	18 custom DCF assumptions (below)	Corresponding camelCase keys (below)
institutional.get_beneficial_ownership	limit	limit
intelligence.get_earnings_calendar	page	page
intelligence.get_dividends_calendar	page	page
intelligence.get_stock_splits_calendar	page	page
intelligence.get_esg_benchmark	year	year
intelligence.get_senate_trading	page, limit	page, limit
intelligence.get_house_disclosure	page, limit	page, limit
alternative.get_crypto_intraday	start_date, end_date	from, to
alternative.get_forex_intraday	start_date, end_date	from, to
alternative.get_commodity_intraday	start_date, end_date	from, to
sec.get_sic_codes	industry_title, sic_code	industryTitle, sicCode

Custom DCF assumptions

Both custom DCF methods expose all 18 assumptions as float | None. Values are forwarded without unit conversion; percentage fields use the units documented by FMP. Omitting an assumption leaves its selection to the provider.

Python argument	FMP query key
revenue_growth_pct	revenueGrowthPct
ebitda_pct	ebitdaPct
depreciation_and_amortization_pct	depreciationAndAmortizationPct
cash_and_short_term_investments_pct	cashAndShortTermInvestmentsPct
receivables_pct	receivablesPct
inventories_pct	inventoriesPct
payable_pct	payablePct
ebit_pct	ebitPct
capital_expenditure_pct	capitalExpenditurePct
operating_cash_flow_pct	operatingCashFlowPct
selling_general_and_administrative_expenses_pct	sellingGeneralAndAdministrativeExpensesPct
tax_rate	taxRate
long_term_growth_rate	longTermGrowthRate
cost_of_debt	costOfDebt
cost_of_equity	costOfEquity
market_risk_premium	marketRiskPremium
beta	beta
risk_free_rate	riskFreeRate

Tool integrations

Existing LangChain and MCP catalog entries expose these optional controls with matching types and hints. The catalog inventory is unchanged: the screener, available-exchanges, and company earnings convenience methods remain SDK-only; applications can wrap their public methods explicitly. The intelligence earnings tool already exposed include_report_times.

market.search_exchange_variants(query) retains its public argument name, but now sends FMP's required symbol query key. For example, pass "MSFT" to search Microsoft's exchange variants. The old query=MSFT wire request returned Apple variants in the live audit.

Provider qualification

Exposing a filter is separate from verifying the returned dataset. The audit found overlapping calendar pages, an empty date-bounded GDP response despite an in-range row in the unbounded series, and unresolved SEC profile identifier behavior. NYSE 2027/2028 holiday and early-close records also have known provider-side defects. No calendar flags, provider payloads, or date boundaries are rewritten. An empty successful response does not establish completeness.

See issue #396 for the provider evidence and the closed connector scope. Correcting or completing these datasets is FMP's responsibility; this release does not qualify a trading-session calendar.