Trados Cloud API - Machine Translation - target segments empty

Hello,

Machine Translation output is not added in the target language segments using Trados Cloud API.

The project is created with success, no error. The machine translation step runs. But when we open the Project online, the machine translation output is not inserted in the target language segments.The target segments stay empty (or source is copied into the target).

The translation engine (DeepL)  is well defined in the template and the machine translation step in the workflow too.

Also, when creating the project online without API, the machine translation outpout is well inserted in the target segment.

Could you please help us here? 

Thank you in advance.

Best,

Brigitte

{
      "id":"6830623d4674622f6fdce787",
      "status":"skipped",
      "taskType":
      {
        "id":"65008e68af67012db01785a1",
        "key":"apply-machine-translation",
        "name":"Machine Translation",
        "automatic":true,
        "scope":"file"
      },
      "input":
      {
        "type":"targetFile",
        "targetFile":
        {
          "id":"6830623b872f877697949d55",
          "name":"file.docx",
          "languageDirection":
          {
            "sourceLanguage":
            {
              "languageCode":"en-GB",
              "englishName":"English (United Kingdom)"
            },
            "targetLanguage":
            {
              "languageCode":"fr-FR",
              "englishName":"French (France)"
            },
            "id":"683062306462aa45c4d61d20"
          }
        }
      },
      "inputFiles":
      [

      ],
      "assignees":
      [

      ],
      "createdAt":"2025-05-23T11:55:40.874Z",
      "failedTask":
      {
        "errors":
        [
          {
            "name":"engineId",
            "code":"empty",
            "value":"en_fr_NEURAL_True"
          }
        ]
      }
    }

Parents Reply
  •   

    I mean, with older versions.  It is related to the massive improvements we have seen with NMT and how workflows have adapted as a result.  So at some point we switched to have all machine translation providers use NMT (as nobody uses statistical any more) and give the results that status.  This also allowed us to remove the confusion over where AT results came from since they are often generated from the TM itself and not machine translation at all.

    Using NMT as opposed to AT was definitely prior to 2024.  With regard to the analysis change I can only find reference to 10 December 2024 (CU13 release) onward, where GroupShare began reporting NMT in its own category rather than under AT/New.  The changes in Trados Studio were more under the hood and made over a number of releases starting towards the end of 2021.

    Paul Filkin | RWS Group

    ________________________
    Design your own training!

    You've done the courses and still need to go a little further, or still not clear? 
    Tell us what you need in our Community Solutions Hub

Children
No Data