Compare
Transept vs ChatGPT for translation
No cardFirst 5 pages freeYour work never trains AI
Literess, the AI editor in Transept
You: Make Grandma grumpier.
Literess: Done. She doesn’t just say it anymore. She grumbles it. sagte → raunzte
Ask in plain words. You approve every change. Meet Literess →
ChatGPT, Claude, and Gemini are remarkable generalists. They write code, summarize papers, draft emails, brainstorm pitches, and yes, they translate. For the occasional translation, a chat window is enough, and ChatGPT now stretches further: a project keeps your files and instructions between chats, and its Word add-in works inside the document you have open. When translation is a recurring job (a translator’s client work, a writer’s weekly newsletter, a marketing team’s launches), you need more than a place to keep context: a glossary applied to every passage, a choice of wording for each sentence, a review step colleagues and clients can join, and a file that comes back in its original layout. Transept is built for the point where “ChatGPT can translate” stops being enough and you need a translation workspace.
Chat window vs translation workspace
| Feature | ChatGPT | Transept |
|---|---|---|
| Single-sentence quality | Yes | Yes |
| Project state across sessions | YesProjects keep chats and files | Yes |
| Glossary enforcement | PartialAs context in a project | YesEnforced on every run |
| Reusable styleguide | YesProject instructions | YesSaved, versioned |
| Document upload & round-trip | PartialWord add-in; layouts may need fixing | YesDOCX/MD/Notion intact |
| Sentence-level alternatives panel | No | Yes |
| Translation memory | NoChat memory, not a TM | YesBuilt in, feeds the AI |
| Smart Proofread | No | Yes |
| Team review & comments | PartialShared projects | Yes |
| Batch: whole documents, many languages | No | YesLite and up |
| Notion & Google Drive | Partial | Yes |
| Audit trail of every change | No | Yes |
| Starting price (paid) | ChatGPT Plus, US$20/mo: a general subscription, not translation volume. A cheaper Go tier sits below it; its price varies by region | €9/$10 a month, excl. VAT (25,000 credits) |
ChatGPT prices checked on October 5, 2026 against its published pricing. ChatGPT sets its own plans and prices, and they may have changed since. View ChatGPT pricing
ChatGPT is great for prompts, and its projects now keep files and instructions from one chat to the next. Transept is built for translation projects: it enforces a glossary on every run, offers alternatives for each sentence, runs QA, and exports DOCX with the formatting intact.
Where the chat window runs out
After the first chat
ChatGPT’s memory and projects can carry your terminology from one chat to the next, as context the model reads. Transept saves your glossary, styleguide, and project, and every translation follows them.
When the layout matters
ChatGPT can now work inside a Word file, though OpenAI notes that complex formatting, tables, and charts may need manual adjustment. Transept works on the file itself: DOCX in, DOCX out, with the formatting intact.
When more than one person is involved
A shared ChatGPT project lets several people chat over the same files and instructions. Transept adds the review layer: comments, team review, client review links, and an audit trail.
A record of who wrote what
Transept counts, by word, how much of a translation the AI wrote and how much you wrote or changed, and lists every edit and approval. Download it as an originality report when a client or publisher asks how the text was made.
~1,000 calls measured
We logged the reasoning of Gemini 3, our engine at the time, on roughly a thousand real translation calls. On plain text the pricier Pro model thought about half as much as Flash. The harder the sentence, the more any model reasons, whatever its tier.
What we measuredTransept runs on the same class of frontier model you use in tools like ChatGPT (currently Anthropic’s Claude), so the model itself is rarely the difference. What matters is everything around it: the glossary added to every prompt automatically, the styleguide that follows you from document to document, the alternatives panel that lets you pick a sentence instead of regenerating it, the QA pass that catches drift after twenty pages, and the export that comes back as a real DOCX with the original formatting. ChatGPT’s projects now keep files and instructions between chats, which covers part of this. The rest is translation-specific, and a general-purpose assistant leaves it to you to ask for each time.
If you translate now and then (a paragraph, an email, a message to a client abroad), keep using ChatGPT. It’s the right tool for that. If translation is a regular part of your work, with several documents a week, clients who each have their own terminology, and deliverables in formats someone else specified, that’s where the chat window breaks down. Moving that work into a tool built for it takes five minutes to set up and pays for itself in the first week.
— Don’t take our word for it —
The editor, in miniature
A working slice of the real app: Literess, the glossary, the styleguide, workflows, and translation memory are all live. Click around.
Simple pricing
1 credit = 1 source word. Pro mode costs 3 credits per word. What each action costs
Free
Always freeFor your first projects
Unlimited users
Translated and proofread on Standard. About five with Pro.
- Pro mode: the model thinks each passage through, 3 credits per source word
- Standard quality: fast, accurate, faithful
- Translation memory: your past translations, reused to keep new ones consistent
- Glossary, styleguide, QA, import / export: all included
- Notion, Drive, TMX / XLIFF: import, export, and bring your own memory
- 68 languages: from Spanish to Tagalog
- Paragraph-by-paragraph translation in every mode
- Literess, your AI literary editor: 10 messages free, then credits
- Originality reports: 2 included
- No card required
Lite
The pocket edition
Unlimited users
- Whole pages and documents in one run
- Unlimited originality reports
- Workflows: chain translation, proofreading, and checks into one run
- API, MCP & webhooks: automate runs, or let AI assistants work in your account
- Teams: create as many as you need and invite anyone for free
- Top up anytime: pack credits never expire
Starter
Most popularFor regular client work
Unlimited users
- 3× Lite’s credits: for weekly work
- Teams: invite anyone for free; add €29/$33 seats to fund a shared credit pool (80,000 credits a month each)
Pro
For higher volumeFor agencies & large catalogs
Unlimited users
- 3× Starter’s credits: headroom for batch work and full catalogs
- Priority support: a direct line and faster replies
- Predictable bill: most months need no extra credit packs
FAQ
- Transept runs on frontier models, currently Anthropic’s Claude, on an engine that isn’t tied to one provider. What it adds is everything around the model call: glossary enforcement, document context, sentence alternatives, QA, review, and export.
- Not directly. Transept pays the model providers itself and includes model access in your plan price, so you don’t need separate API keys or subscriptions.
- Import your last project, paste in the glossary you’ve been carrying in your prompt, and save a styleguide from one of your best translated documents. The setup takes 10 minutes, and you save time from the next document on.
- The first draft is comparable, because Transept runs on the same class of frontier models (currently Anthropic’s Claude). The difference comes after it: Transept enforces your glossary on every run and runs a QA pass, steps that in a chat happen only when you ask for them.
- For a sentence or a paragraph, yes. ChatGPT (like Claude and Gemini) translates fluently, often better than a phrase-book tool, and its quality holds up on short, self-contained text. Long documents ask more of it: the model reads your glossary as context rather than checking each term, and OpenAI notes that complex formatting, tables, and charts may need manual adjustment in its Word add-in. Transept runs on the same class of frontier model but adds your glossary to every call, keeps one styleguide across documents, and exports a real DOCX, so the quality holds for the whole document.
- It depends on the length and the setup. Under a few thousand words, ChatGPT translation quality is strong. On longer work, a ChatGPT project helps: it keeps your files and instructions, so your terminology travels from chat to chat, but the model applies it as context, and OpenAI notes that complex formatting and tables may need manual adjustment in its Word add-in. A translation workspace closes that gap, with deterministic glossary enforcement, a QA pass (Smart Proofread) that re-reads for drift, and a formatted export.
- ChatGPT does more with documents than it used to. It reads uploaded DOCX and PDF files, its Word add-in (on every plan, including Free) works inside the document you have open, and ChatGPT Work can create or edit documents, including Google Docs. OpenAI’s own help notes that complex formatting, tables, and charts may need manual adjustment. Transept splits the document into blocks, translates each one with the whole document as context, and puts a Word file back together with its original formatting intact. A PDF can be translated onto its own pages or as text only.
- Literess, the assistant inside Transept, works on the document you have open. Ask her to rework a passage, build a glossary from a chapter, or proofread and fix what she finds. She reads your glossary and styleguide and searches your translation memory, so she answers with your project in front of her. Questions about the app are free, work on your translation costs the credits her answer takes, and changes to your documents wait for your approval. On Free, your first 10 messages are free.
- ChatGPT’s projects keep chats, files, and instructions together across conversations, and you can share one with collaborators, which helps. What they don’t add are translation tools: a glossary checked on every translation, an alternatives panel for each sentence, and a QA pass. On a long document, those gaps add up.
- For purely programmatic translation, the OpenAI API works fine. Transept is worth switching to when you want a ready-made workspace (alternatives, glossary management, QA, comments, exports) instead of building those parts yourself.
- Transept’s engine isn’t tied to one AI provider, so it can adopt new frontier models as they come out without breaking your existing workflows. When a stronger model proves itself, it can power the modes you already use.
Keep reading around this comparison
- Use caseTranslate your book with AI without losing the voice→A character glossary that holds for 300 pages, a styleguide built from your own chapter, and alternatives for every line of dialogue. For…
- FeatureA glossary the AI actually follows→Character names, product terms, brand phrases, client vocabulary. Build a glossary once (by hand, from a file, or from a sample document)…
- CompareTransept vs DeepL for long documents→DeepL is the default for fast, high-quality AI translation, and Transept doesn’t claim to beat it on a single sentence. Most translation…
- GuideWhat It Costs to Translate a Novel in 2026: AI vs. AI + Editor vs. Human→What translating a book costs in 2026, three ways: a human literary translator, an AI draft with a human editor, and AI alone. Real rates…
- GuideThe best AI tools to translate a novel in 2026→Eleven ways to translate a novel with AI in 2026, including Kindle Translate, DeepL, ChatGPT, Claude and the tools built for books. What…
Built for work you answer for
Private by default
Your text is never used to train AI, not by us and not by any provider we use, and it stays out of our logs. Delete your account and your text is deleted with it.
How we handle dataSecurity you can check
Hosted in Germany, in Hetzner’s ISO 27001-certified data center. Every sub-processor is listed, and a DPA is one email away.
All sub-processorsGive it context
Glossaries, styleguides, translation memory and per-string notes go with every request. You set them up once per project.
GlossariesYou’re the human in the loop
The AI drafts; you decide. Review every change side by side, keep or revert it, and send a review link when a client needs to look.
Review linksAPI, MCP and webhooks
The full API on every paid plan. Connect Claude or another AI assistant over MCP, or start runs from your own tools.
Developer guideSomething bigger?
Large catalogs, custom workflows, a DPA for procurement. Transept stays self-serve; write to us for the rest.
support@transept.aiFeatured on
