# Long-context translation leaderboard

Multilingual benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelchoice.dev/leaderboards/benchmarks/multilingual-long-context-translation

Long-context translation evaluation (Cohere). Every model the catalog carries a published Long-context translation value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Multilingual |
| Measure | xCOMET-XL quality score |
| Tasks | Two book chapters per call |
| Difficulty | Long-document translation |
| Published by | [Introducing North Small Translate](https://cohere.com/blog/north-small-translate) |

Cohere's long-document translation evaluation: two book chapters translated in a single call, scored per paragraph with xCOMET-XL.

## Long-context translation ranking

1 model with a published Long-context translation value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | xCOMET-XL quality score | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [North Small Translate](https://modelchoice.dev/models/north-small-translate-1-0) | Cohere | 48.9 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
