Long-context translation evaluation (Cohere) (Long-context translation)
We show this table for reference; we do not rank on it.
Cohere's long-document translation evaluation: two book chapters translated in a single call, scored per paragraph with xCOMET-XL.
Benchmark score on Long-context translation — November 13, 2026
We compile the Long-context translation rows from provider self-reports. North Small Translate leads the table at 48.9%. We do not use these results to rank models overall.
1 modelMultilingualCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (1 model)
The one result on this page was reported by the lab that built the model.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | North Small TranslateCohereReported by Cohere | 48.9% | Reported by Cohere | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
About Long-context translation
Year
2026
Tasks
Two book chapters per call
Format
xCOMET-XL quality score
Difficulty
Long-document translation
A provider-designed evaluation published with the North Small Translate launch. Quality is measured for each paragraph in isolation via xCOMET-XL after translating two chapters of a book in one call. BenchLM keeps it display-only because the source texts and scoring pipeline are not public.
Freshness and provenance
Version
Long-context translation 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does Long-context translation measure?
Cohere's long-document translation evaluation: two book chapters translated in a single call, scored per paragraph with xCOMET-XL.
Which model scores highest on Long-context translation?
North Small Translate by Cohere currently leads with a score of 48.9% on Long-context translation.
How many models are evaluated on Long-context translation?
1 AI models have published results on Long-context translation in the BenchLM catalog.
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.