OANDA Exchange Rates

OANDA Exchange Rates is a currency data API backed by OANDA and central bank sources. Agents fetch spot rates and candles, pull historical ranges, convert amounts, list currencies, and check remaining plan quota.

Verified 6 actions API key available

OANDA Exchange Rates website Platform Documentation Capability data verified 2026-08-25

A scheduled restatement job starts for a batch of foreign currency invoices

Get Remaining Quotes reports the quota left on the plan before anything is quoted

List Currencies confirms which codes the chosen data set actually quotes

The agent counts the distinct currency and timestamp pairs the batch genuinely needs

Convert Amount prices one sample invoice and returns the rate, the side, and the timestamp it used

The sample conversion and the quota arithmetic post to the finance channel

The finance lead decides whether to run the full batch, narrow it, or wait for the next quota reset

What This Integration Enables

OANDA sells two unrelated products under one brand, and this connector is the one finance teams want rather than the one traders do. The Exchange Rates API answers a narrower question than a trading feed: what was this pair worth at this moment, on this named price source, and on which side of the quote. You can point it at OANDA's own feed, at the cryptocurrency feed, or at a specific central bank's published list, and the answer changes accordingly. That choice is the product. A rate is only defensible when you can say where it came from.

FlowRunner agents read spot quotes with bid, ask, and midpoint, pull a candle at a moment, fetch a date range as JSON, CSV, or XML, list which currency codes a data set can actually quote, and convert an amount with the side named explicitly. The connector also handles two vendor rules that quietly break naive flows: timestamps must land on a quarter hour boundary, and only one of the two live API generations accepts a start and end date. Both are handled inside the connectors layer rather than left as a debugging exercise. And because the API bills per quote, the quota read is free and worth running first.

Without FlowRunner

Rates with no provenance A workbook carries a converted figure and no record of the source, the side, or the moment

Quota spent without warning A loop asks for every currency at once and the plan runs dry partway through the period

The wrong OANDA key A trading token is pasted into a rates integration, and the error names a format rule rather than the cause

With FlowRunner

Provenance on every figure Convert Amount reports the rate it read, the side it used, and the timestamp it priced at

The meter is read before it is spent The free quota check runs first, and a batch that would drain the plan waits for a person

Endpoint choice handled for you Range queries route to the generation that accepts a start and end, and timestamps snap to the boundary the API demands

Use Case Scenarios

Invoicing in the customer's currency, priced on a stated side

A subscription renews in Stripe in one currency while the ledger reports in another. Before the invoice posts to QuickBooks Online, the agent calls Convert Amount with the side the billing policy names, usually midpoint for reporting or ask for a figure the customer is actually charged. The response carries the rate, the side, and the timestamp, and all three are written onto the invoice memo. When a customer later disputes the converted total, the answer is on the document rather than in somebody's memory.
A rate series that feeds a model instead of a screenshot

An analyst needs a stretch of history for a hedging model. The agent calls Get Historical Rates for the range, sets the output format to CSV, and lands the file straight into Google Sheets or a table in Snowflake. Get Candle fills in the open, close, high, and low at specific decision points the model cares about. The series arrives already shaped, and the same job can be re-run for a different data set to test whether the conclusion depends on the price source.
A rate move that reaches the pricing owner, not a dashboard

The API publishes no webhooks, so the agent polls Get Spot Rate on a schedule for the two or three pairs the business actually prices in, rather than asking for everything. When a pair crosses a threshold held in Airtable, the agent assembles the current bid, ask, and midpoint and posts them to the pricing owner in Slack with the affected price list attached. Nobody is asked to watch a chart, and nobody discovers the move a week later in a margin report.

Human-in-Loop Highlight

This API bills per quote, and the quota is a single meter shared by every flow on the key. That makes a large backfill a spending decision, not a technical one, because quota consumed by a restatement run is quota the nightly billing conversion will not have. Get Remaining Quotes costs nothing, so the agent reads the meter before it starts. When the batch in front of it would take more than the plan has left, it stops and posts to the finance lead: "This restatement needs quotes for eleven currencies across sixty timestamps. The plan's remaining balance will not cover that and the nightly billing conversion. Run it in full, narrow it to the three currencies we invoice in, or defer to the next reset?" The lead answers. That is the difference between an automation exception handled by a person and a billing flow that fails silently at the end of the month because a backfill got there first.

Agent processes routinely

Detects exception requiring judgment

Clear match Continues automatically

Ambiguous Routes to human via preferred channel

Human decides

Agent resumes with decision

Agent Capabilities

6 actions

Rates

Get Spot Rate Returns bid, ask, and midpoint for a base currency against the quote currencies you name. Leaving the quote currencies empty asks for every currency at once, which is the fastest way to spend quota, so name the ones you need.
Get Candle Returns open, close, high, low, and averages at a single moment. Used when a workflow needs the shape of the market at a decision point rather than a single price.
Get Historical Rates Returns a start to end date range, and it is the only operation that accepts one. Output can be JSON, CSV, or XML, which makes it the right call when the destination is a spreadsheet or an archive rather than a branch in a flow.

Reference

List Currencies Returns the currency codes the plan can quote for a given data set. The authoritative answer for what the base and quote parameters will accept, since the OANDA feed, the cryptocurrency feed, and a central bank list each carry different currencies.

Account

Get Remaining Quotes Reports how much quota is left on the plan, and costs nothing to call. Used before any loop or backfill so a batch cannot quietly exhaust the meter every other flow depends on.

Conversion

Convert Amount Converts an amount between two currencies and reports the exact rate, side, and timestamp it used. The side is an explicit choice: midpoint for reporting, bid or ask for a figure that reflects what a real trade would cost.

Frequently Asked Questions

What can FlowRunner do with OANDA Exchange Rates?

FlowRunner agents can run Get Spot Rate, Get Candle, and Get Historical Rates in OANDA Exchange Rates, plus 3 more actions.

Does connecting OANDA Exchange Rates to FlowRunner require OAuth?

No. OANDA Exchange Rates connects to FlowRunner with an API key, no OAuth flow required.

Can OANDA Exchange Rates trigger a FlowRunner workflow automatically?

OANDA Exchange Rates doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.

Start building with OANDA Exchange Rates

Free plan, no card required. Connect in minutes.

Start Building Free Talk to us