Machine translation produces volume. Post-editing determines whether that volume is usable. We offer both light and full MTPE at published rates, clearly distinguished from human translation.
What is included
MT output assessment before pricing
Light or full post-editing as agreed
MTPE specialist assigned
Track-changes file for review
Error summary delivered
Per-1,000-word pricing, no minimum surprises
5 business days for standard MTPE projects.
What we mean by light and full post-editing
Light post-editing (LPE) corrects fluency errors, obvious mistranslations, and formatting issues to produce content that is comprehensible and sufficiently accurate for its intended use, typically internal documents, low-priority content, or content with a short lifecycle. It does not aim to reach human translation quality.
Full post-editing (FPE) brings MT output to a quality equivalent to human translation. Every sentence is reviewed against the source text, terminology is verified, and style is corrected to match the agreed standard. This is appropriate for public-facing content, regulatory documents, or anything where quality matters.
What machine translation we work with
We work with output from DeepL, Google Translate, and most major MT systems. If you have a custom MT engine, we assess the output quality before quoting. MT quality varies significantly by language pair and domain, and our pricing reflects the actual post-editing effort required, not a flat-rate assumption.
Transparency about MTPE
We quote MTPE and human translation as separate services at separate prices. If a client is unsure which is appropriate for their content, we assess the MT output and recommend. We do not deliver MTPE as if it were human translation.
FAQ for this service
For most general content, yes. For highly technical, legal, or regulated content, the post-editing effort required for acceptable quality often makes human translation more cost-effective. We assess and advise.
We do not deliver raw MT output as a service. If you want MT output only, you can access that directly from DeepL or similar. What we deliver is post-edited MT that has been reviewed against the source by a professional.
