How TravelTime pricing works

TravelTime is a fixed annual subscription.

Two things set the price: the capacity your application needs, and the service level you need from us. We don't count requests, locations, elements or results, and there are no overage charges.

Developer plans are free forever, and testing at scale costs nothing. Production plans are for commercial use. Discover the plans here.

Everything below explains how the model works in practice. If you want the reasoning behind it rather than the mechanics, start with our pricing model.

What are the two factors that set the price?

Capacity

Capacity is provisioned throughput, measured in hits per minute (HPM). We agree the peak throughput your workload needs and provision it for your subscription term. Inside that capacity, your usage can move freely.

Service level

This is the assurance you need around TravelTime: the level of support you want from the team that built it, SLAs, and enterprise controls such as single sign-on and multi-factor authentication. Our security and compliance documentation lives in the Trust Center.

That's it. No line item for volume, no per-endpoint or per-element pricing, no tiering by feature.

Why doesn't TravelTime price on usage?

We don't price by API usage because usage-based pricing shapes what gets built.

When every call has a price attached, the meter starts making product decisions. Teams cache less than they should because caching is prohibited, or they can't refresh results as often as the data deserves. They cut the number of destinations in a search to keep the element count down. They shelve the continuous optimisation, the model training run, the nightly recalculation of a whole territory, because the maths doesn't work at the unit price.

We've written about how that becomes a form of technical lock-in rather than just a budget problem.

It shows up most clearly in the comparisons. Our breakdowns of Distance Matrix API pricing and Isochrone API pricing show what per-element billing does to a real workload as it scales. And if you're currently on Google, the Google Maps API comparison sets out where the restrictions bite hardest.

We don't meter requests, and we don't intend to. The point of the model is that the cost of your next calculation is zero, so the decision about whether to run it is a product decision rather than a budget one.

What happens if I exceed my capacity?

If you exceed provisioned capacity, you get a 429 and you can retry. Nothing is billed as overage and nothing is lost. At peak you may wait longer for a result, but you'll always get it.

If your capacity needs to change, we agree a new configuration before the price changes.

There is no mechanism by which a busy month turns into a bigger invoice. Live throughput and incident history are on our status page.

What does the subscription actually pay for?

Your TravelTime subscription buys access to our data, modelling and compute, plus the rights to put TravelTime inside your product.

• Accurate multimodal transit models: Proprietary transport data and models for traffic-aware driving, full-schedule public transport, cycling and walking, across every endpoint, globally. You can check our numbers against other providers in the accuracy benchmarking tool, and we published the full methodology behind it.
• High performance at huge scale: Infrastructure built to calculate enormous numbers of travel times quickly and consistently. See the response time benchmarks.
• Production reliability and support: Service, support and operational assurance for a dependable integration.
• Contractual data rights and caching: Rights to use, cache, retain, refresh and reuse results in the way your product and architecture need, documented in your contract.

Every customer gets every endpoint, dataset and model. We don't sell partial access, and there's no discount for using less of it. That includes the new Data API (POI data, geocoding, map tiles, environmental data, and demographic and population data) at no extra cost.

Accuracy doesn't vary by plan either. Every key type gets our best model quality, including free developer keys.

Is this like buying cloud?

Provisioned capacity is a familiar model to anyone who has bought compute. You size the throughput you need, you keep it available, and the workload moves around inside it. It could be spiky through the day, heavier at some points in the week, and quieter overnight.

The alternative is metering, where a normal peak in demand becomes a variable line on an invoice you can't forecast. With provisioned capacity you know the ceiling you're working with from the first day of the term, and the peaks are already paid for.

Is testing really free?

Yes, testing is free. And it's uncapped and untimed.

Free developer keys run at 120 HPM with up to 10,000 requests each for Geocoding, Map Tiles, POIs, Environmental Data and Demographic Data. They never expire and they don't need a credit card. If you'd rather try the endpoints before writing any code, try the API playground.

For an evaluation, we'll provision unlimited access at any scale for as long as you need. A long-term partnership only works if you've proven the accuracy and performance first, so we're happy for you to test properly rather than take our word for it.

The one boundary: free keys are for research, development and evaluation, not commercial use. If TravelTime is part of a process or product that makes money, that needs a Production subscription — that includes internal commercial use, not just customer-facing products.

Who is TravelTime not for?

TravelTime isn't always competitive if you make a low volume of requests.

Customers come to us when per-request billing has become the constraint on what they can build. That's the problem the TravelTime model solves, and it tends to surface in a few places:

• Search and match: where every user query e.g. a property or hotel search is a live call
• Spatial analysis: where a single study can generate billions of results
• Routing and fleet management: where optimisation means calling the same API repeatedly
• AI and LLM tooling: where an agent's call volume is unpredictable

Our case studies show what teams built once the meter stopped being a factor.

How do we get to a price?

Step 1: Tell us what you're building. A short conversation about your use case and the shape of your usage.

Step 2: Test for free. Full API and support access, at whatever scale your evaluation needs.

Step 3: We agree a fixed price. Sized to your capacity and service level, with room to grow.

Step 4: Build without compromise. Unlimited calls, full caching rights, and a support contact when you need one.

Subscriptions run for a minimum of 12 months, and we bill in GBP, USD, EUR, AUD and NZD.

See the full plan breakdown on our pricing page, get started with a free key, or talk to our team about what your fixed price would be.