Currency Exchange

Currency Exchange — Norges Bank exchange rate MCP server for Claude, Cursor and any MCP client

Currency gives an AI assistant Norges Bank's daily reference rates as structured data: the rate between any two of 38 currencies on any date back to about 1980, conversion of an amount, a daily series for charting — and World Bank buying-power multipliers for pricing one list price into many markets. It runs as a hosted MCP server over HTTPS.

Connect

{ "mcpServers": { "currency": { "type": "http", "url": "https://currency.publifye.com/mcp" } } }

Claude Code: claude mcp add --transport http currency https://currency.publifye.com/mcp, then /mcp to sign in.

The OAuth flow runs in the browser on first use, with a Publifye account. Access is granted through that sign-in only; there is no API key to create, copy or rotate. The discovery chain is the same as for the other servers: [../../docs/connect.md](../../docs/connect.md).

Where the rates come from

The rates are Norges Bank's own daily reference rates, quoted against the Norwegian krone. This service syncs and serves them. It does not compute, estimate or interpolate a rate.

That last sentence is the one that matters when an AI is asking. Norges Bank publishes once per business day, at about 16:00 CET, so weekends, Norwegian public holidays and any day the bank did not publish simply have no rate. A date with no published rate returns the series as published rather than an invented value — no carry-forward, no interpolation, no nearest-neighbour. A model that wants a number for a Sunday has to decide for itself what to do about it, which is the correct place for that decision to be made.

Cross-rates — USD→EUR, say — are derived through NOK, because NOK is the only pair Norges Bank quotes. The derivation is arithmetic on two published rates, not a third source.

The buying-power tools return two numbers, and the difference matters

get_buying_power and list_buying_power exist for one job: pricing a USD list price into a local market. They return `factor` and `raw_factor`, and they are not the same thing.

• `factor` is the recommended multiplier. It is never above 1.0 — the USD price is the maximum anyone pays regardless of how wealthy the local market is — and it is floored at 0.35 because payment fees make smaller charges uneconomic.
• `raw_factor` is the honest, uncapped measurement: GNI per capita (PPP) relative to the USA, square-root compressed. It can exceed 1.0. Norway measures about 1.10.

Applying raw_factor to a Norwegian price charges about 10% above list. That is exactly the mistake the cap exists to prevent, which is why both numbers are returned with capped and floored flags saying whether policy moved one away from the other. If in doubt, use factor.

An unknown country, missing data or an upstream outage fails safe: factor 1.0, with unknown or fallback set and the reason in note — never a silent guess.

What the tools do

Every tool is documented with its exact description, annotations and input schema — 7 in all, generated from the service's own registry, never written by hand.

Machine-readable: [tools.json](tools.json) carries all 7 callable tools with full JSON Schema, plus every excluded bucket listed by name so the count is auditable.

All seven are read-only and idempotent. None is marked open-world: every answer is served from data this service has already synced, so the same call on the same day gives the same answer.

What it does not do

• No live or intraday rates. One publication per business day, about 16:00 CET. A day's rates are stable once published, and therefore cacheable.
• No currency Norges Bank does not quote. 38 plus NOK, and no more.
• No invented values. No interpolation across a gap, no carry-forward into a weekend.
• No trading, no execution, no advice. These are reference rates, not dealable prices, and a reference rate is not what you will be charged by a bank or a card network.
• Not a full annual-accounts or economics source. The buying-power figures are one World Bank indicator, not a model of an economy.

Data and licence

Exchange rates are published by Norges Bank. Buying-power figures are World Bank GNI per capita (PPP), released annually, around July. Results carry reference_year — the World Bank's release year — separately from fetched_at, which is when this service last pulled it. Those are not the same thing and conflating them is how a stale figure gets reported as current.

data_origin and fallback say whether a buying-power answer came from a live pull or from the table embedded in the binary, so a degraded answer is never indistinguishable from a fresh one.

Currency Exchange is operated by Publifye AS and is not affiliated with, or endorsed by, Norges Bank or the World Bank.

Plans

A free Publifye account grants a small daily allowance on a trial; Currency Access lifts it to a monthly quota. Plans and prices can change; the current ones are always on [https://currency.publifye.com](https://currency.publifye.com), and where this page and the site differ, the site is right.