Yes. CotranslatorAI makes a very good general-purpose AI assistant as long as you only need self-contained, text-based responses rather than features such as web browsing or file analysis.

The underlying language model contains a substantial amount of knowledge. You can ask how to write an Excel formula, get an explanation of a specific historical event, draft an email, or work through an idea. You don’t have to limit yourself to translation questions.

Using CotranslatorAI with the OpenAI API also gives you stronger default privacy protections than a personal ChatGPT account: OpenAI does not use API inputs and outputs to train its models by default.

Cost depends on how much you use it and which model you choose. API usage is billed separately from your CotranslatorAI license, based on usage rather than a flat monthly chatbot subscription. In addition, when working in CotranslatorAI, you get unlimited access to the latest models; you aren’t throttled in the same way the chatbots do, even the paid versions.

You may also find the answers in CotranslatorAI less verbose than those from a consumer chatbot, depending on the model and your instructions. If you prefer short answers, tell the AI: “Answer directly and briefly. Add detail only if I ask.”

That can help you stay focused and get the information you need without adding unnecessarily to the information overload we already deal with every day.

We haven’t built a single search tool into our website or into CotranslatorAI. That’s because you don’t need to; you can get all those answers easily on your own.

One option is to put The CotranslatorAI Guide into NotebookLM and ask questions about it. NotebookLM can point you to relevant passages in the Guide, making it easier to find instructions for your specific task. You’ll find the Guide here:

The CotranslatorAI website also contains a lot of helpful information you might not notice at first. Browse through the menus to find documentation, tutorials, and other resources.

For an even easier starting point, paste https://cotranslatorai.com/ into a chatbot that can browse the web and ask your question. For example:

“Search this website for instructions on adding terminology and previously translated segments to CotranslatorAI. Explain the workflow and give me links to the relevant resources.”

This can get you answers directly in the chatbot, along with links to explore further.

The AI Survival Kit that comes with the CotranslatorAI Bundle comes with a variety of resources for better workflow technique. You can also browse through the resources at My First 5 Wins.

You can put both terminology pairs and previously translated source and target segments into your prompt. Terminology pairs help the AI use your preferred terms; translated segments give it examples of your project’s translated style and sentence structure.

For terminology, copy the entries you need from your  termbase, or export the relevant terminology and open it in Excel. Then copy the source and target columns into the prompt. Keep each pair together, preferably on its own line, separated by a descriptive symbol or a tab.

Add an instruction such as:

“Use the following terminology when translating this project. Each line contains a source term followed by its approved target-language equivalent.”

For multiple translated segment pairs, ordinary copying from memoQ’s translation grid may not give you both columns in the format you need. There is a workflow for adding multiple segments individually, described in the last section of The CotranslatorAI Guide.

Another workaround I’ve used is ABBYY Screenshot Reader. Capture the relevant source and target rows in your CAT-tool editor, extract the text with OCR, and paste it into the prompt. You can use another OCR tool if you prefer. You don’t need to export the entire bilingual document for this approach. Just capture the rows you want to use as reference. Check the extracted text for recognition errors and make sure each source segment is clearly paired with its translation.

Then add an instruction such as:

“The following source and target segment pairs are approved translations from this project. Use them as reference examples for terminology, style, and sentence structure when translating new segments.”

Another option is to keep an exported bilingual file open alongside your project and copy only the relevant sections into the prompt. Although this involves an initial export, it can be more convenient than repeatedly capturing text from the CAT-tool grid.

Tags can get corrupted when you copy text directly from a CAT tool’s translation editor into CotranslatorAI. For example, in memoQ, you may see “rpr” change to “OBJ.”

We have an AutoHotkey routine that fixes the tags for memoQ. I am happy to provide it for free to users of the CotranslatorAI paid editions.

Alternatively, if you use memoQ, work with an exported RTF bilingual file rather than copying directly from the translation grid. This preserves the tags and avoids the problem. The same workaround does not preserve tags in Trados.