Simple pricing
Our pricing is volume-based, just like machine translation. You only pay for what you use, so costs stay in line with the value you're getting as you scale. Volume is based on the source characters. So you know what the price will be before you start a project, and the price will be the same for every target language. On high-volume enterprise post-editing workflows, customers start saving up to 75% by adding ModelFront.
Talk to us! A minimum commitment is required to train, test and deploy a custom model. One custom model can support all your language pairs. A custom model is strongly recommended for most use cases.
For custom models, we cannot provide a free trial. Custom models are strongly recommended for certain use cases, like hybrid translation. Before you order your custom model, ModelFront will assess your scenario and your data, to share whether or not it is a fit.
Only original source-side text characters are counted, the translated target-side text is not counted. One character is counted the same regardless of whether it is a whitespace character, a decomposed umlaut or a UTF-8 Chinese ideogram. This avoids lossy conversions and makes it easy for client applications to know the exact count. You are only billed for successful responses, not for requests that fail due to bad requests or server errors.
This option is typically only used for development, and not used in translation management system (TMS) integrations. If you choose to use an integrated third-party machine translation option, it is billed at cost. If you use the * option ("Let ModelFront choose for you…"), it is billed at $200 / 1M characters.
Filtering translation memories? Translating billions of characters via the API? Yes, talk to us about a volume discount.