Multi-model AI agents, translation memory, intelligent QA, and subtitle workbench — unified in one platform. Desktop + Web dual access, with full local LLM support for absolute data privacy.
Not just another CAT tool. WeTrans is an AI-native platform where neural models, agents, and human expertise converge.
Connect GPT-4, Claude, DeepSeek, Gemini, or run Ollama locally. Context-aware translation with document-level reasoning.
Automated workflows powered by intelligent agents — review, QA scoring, terminology extraction, all driven by AI.
Waveform timeline, millisecond sync, bilingual preview, ASR speech recognition — cinematic-grade subtitle production.
High-performance fuzzy matching engine with millisecond retrieval across million-entry corpora. Build linguistic assets — never translate the same sentence twice.
AI + rule-based QA: detect omissions, number mismatches, terminology deviations, and CPS/CPL limits automatically.
No cloud uploads required. Connect to local Ollama models for a completely air-gapped, private translation pipeline.
Word, PDF, PPTX, InDesign, SRT, ASS, VTT — parse complex layouts and maintain original formatting on export.
Native high-performance architecture. Load million-word projects in seconds, handle bulk translations without lag — 10x faster than alternatives.
AI capabilities rest beneath a restrained interface. Less noise, more focus on your words.
Side-by-side bilingual view, syntax highlighting, real-time TM matching, keyboard-centric workflow.
Waveform visualization, millisecond timeline tweaking, ASR auto-transcription, perfect audio-visual sync.
Review agent, QA agent, scoring agent — your AI-powered translation quality control center.
Track progress, manage TM/TB assets, split and merge files, coordinate multi-file projects.
Free 15-day trial with all AI features unlocked. No credit card required.
Start free, upgrade when you need more. No credit card required.
New user welcome bonus
Individual translators & light use
Full-time linguists & teams
Offline deployment & data sovereignty
Ollama local models & GLM-4-Flash are always free — 0 credits. · Manage your account →
WeTrans is an offline-first desktop application. All project files, translation memories, and termbases are stored on your local machine. When using local AI models via Ollama, your entire pipeline is 100% physically private — no data ever leaves your computer.
We currently offer a deeply optimized version for Windows. The Rust core engine for macOS is ready and we're finishing UI adaptation — expect a release next quarter.
After the 15-day trial, you won't be able to create new projects or export files, but all your data remains safely on your machine. Purchasing a license key instantly restores full AI-powered functionality.
WeTrans integrates with OpenAI GPT-4o, Anthropic Claude, Google Gemini, DeepSeek, Zhipu GLM, and any model running locally via Ollama. You can switch models per-project or even per-segment.