WMT26 General MT: WMT26 General MT (human evaluation, cESA Avg. Rank)

Configuration: Gemini 3.1 Pro

Notes: WMT26 General MT findings (2026.wmt-1.48), Figure 3: Avg. Rank 4.8 for the system listed as 'Gemini 3.1 Pro', averaged over the 23 language pairs it entered; rank 1 is best, so lower is better (not a 0-100 score). Per-pair cESA scores (0-100, higher is better, not comparable across pairs): cs-vi 88.5, cs-de 84.0, en-cs 75.5, en-de 78.2, en-is 84.9, en-zh-Hant 70.1, en-arz 83.7, en-ru 79.4, en-be 77.6, en-lld 45.6, en-se 61.3, en-et 81.1, en-ja 90.3, en-kk 80.0, en-id 83.5, zh-ja 75.7, en-ko 83.5, en-th 86.2, en-lij 12.4, cs-uk 84.1, en-uk 84.3, en-hy 80.7, en-zh-Hans 79.3. Preview endpoint: the paper names only 'Gemini 3.1 Pro'. Source: WMT26 organizers. Retrieved 2026-10-27.

Configuration: GPT 5.5

Notes: WMT26 General MT findings (2026.wmt-1.48), Figure 3: Avg. Rank 8.6 for the system listed as 'GPT 5.5', averaged over the 23 language pairs it entered; rank 1 is best, so lower is better (not a 0-100 score). Per-pair cESA scores (0-100, higher is better, not comparable across pairs): cs-vi 89.5, cs-de 84.8, en-cs 77.1, en-de 77.1, en-is 79.7, en-zh-Hant 68.6, en-arz 80.8, en-ru 76.0, en-be 69.6, en-lld 29.7, en-se 49.3, en-et 80.3, en-ja 86.1, en-kk 81.9, en-id 80.3, zh-ja 75.3, en-ko 80.9, en-th 82.0, en-lij 38.3, cs-uk 85.8, en-uk 80.9, en-hy 80.0, en-zh-Hans 73.3. Source: WMT26 organizers. Retrieved 2026-10-27.

Configuration: DeepSeek V4 Pro

Notes: WMT26 General MT findings (2026.wmt-1.48), Figure 3: Avg. Rank 17.5 for the system listed as 'DeepSeek V4 Pro', averaged over the 23 language pairs it entered; rank 1 is best, so lower is better (not a 0-100 score). Per-pair cESA scores (0-100, higher is better, not comparable across pairs): cs-vi 77.0, cs-de 81.3, en-cs 56.6, en-de 66.7, en-is 71.4, en-zh-Hant 64.2, en-arz 80.0, en-ru 66.2, en-be 63.0, en-lld 22.9, en-se 43.8, en-et 71.8, en-ja 83.2, en-kk 74.3, en-id 78.9, zh-ja 72.8, en-ko 74.4, en-th 71.9, en-lij 20.1, cs-uk 82.2, en-uk 80.4, en-hy 67.3, en-zh-Hans 75.3. Source: WMT26 organizers. Retrieved 2026-10-27.

Configuration: Gemma 4 - 31B

Notes: WMT26 General MT findings (2026.wmt-1.48), Figure 3: Avg. Rank 22.0 for the system listed as 'Gemma 4 - 31B', averaged over the 23 language pairs it entered; rank 1 is best, so lower is better (not a 0-100 score). Per-pair cESA scores (0-100, higher is better, not comparable across pairs): cs-vi 75.7, cs-de 76.9, en-cs 60.1, en-de 67.8, en-is 43.6, en-zh-Hant 74.3, en-arz 80.5, en-ru 73.7, en-be 49.9, en-lld 3.6, en-se 4.7, en-et 45.8, en-ja 80.5, en-kk 64.4, en-id 75.7, zh-ja 72.2, en-ko 76.2, en-th 80.0, en-lij 6.6, cs-uk 79.3, en-uk 76.3, en-hy 44.0, en-zh-Hans 72.3. WMT26 names 'Gemma 4 - 31B'; mapped to the instruction-tuned slug. Source: WMT26 organizers. Retrieved 2026-10-27.

Configuration: Command A+

Notes: WMT26 General MT findings (2026.wmt-1.48), Figure 3: Avg. Rank 26.0 for the system listed as 'Command A+', averaged over the 23 language pairs it entered; rank 1 is best, so lower is better (not a 0-100 score). Per-pair cESA scores (0-100, higher is better, not comparable across pairs): cs-vi 73.2, cs-de 71.0, en-cs 61.3, en-de 70.8, en-is 42.0, en-zh-Hant 61.9, en-arz 80.1, en-ru 70.4, en-be 53.7, en-lld 2.9, en-se 2.6, en-et 56.6, en-ja 81.4, en-kk 57.8, en-id 76.8, zh-ja 66.9, en-ko 63.0, en-th 70.1, en-lij 7.0, cs-uk 76.9, en-uk 75.9, en-hy 12.5, en-zh-Hans 65.8. Organizer baseline 'Command A+', not the Cohere CAT+ submission. Source: WMT26 organizers. Retrieved 2026-10-27.

Configuration: Mistral Medium 3.5

Notes: WMT26 General MT findings (2026.wmt-1.48), Figure 3: Avg. Rank 29.3 for the system listed as 'Mistral Medium 3.5', averaged over the 23 language pairs it entered; rank 1 is best, so lower is better (not a 0-100 score). Per-pair cESA scores (0-100, higher is better, not comparable across pairs): cs-vi 79.9, cs-de 72.5, en-cs 67.8, en-de 65.5, en-is 25.7, en-zh-Hant 55.5, en-arz 70.6, en-ru 70.1, en-be 43.2, en-lld 2.9, en-se 2.3, en-et 44.0, en-ja 74.1, en-kk 48.8, en-id 65.0, zh-ja 56.5, en-ko 58.3, en-th 53.6, en-lij 6.0, cs-uk 78.1, en-uk 70.3, en-hy 46.3, en-zh-Hans 67.6. Source: WMT26 organizers. Retrieved 2026-10-27.

Configuration: GPT OSS 120B

Notes: WMT26 General MT findings (2026.wmt-1.48), Figure 3: Avg. Rank 32.1 for the system listed as 'GPT OSS 120B', averaged over the 23 language pairs it entered; rank 1 is best, so lower is better (not a 0-100 score). Per-pair cESA scores (0-100, higher is better, not comparable across pairs): cs-vi 66.3, cs-de 62.9, en-cs 52.1, en-de 58.5, en-is 46.6, en-zh-Hant 58.1, en-arz 77.6, en-ru 56.6, en-be 30.5, en-lld 5.5, en-se 2.3, en-et 47.2, en-ja 78.2, en-kk 48.7, en-id 68.3, zh-ja 64.9, en-ko 54.6, en-th 45.5, en-lij 4.4, cs-uk 63.1, en-uk 58.1, en-hy 32.1, en-zh-Hans 61.9. Source: WMT26 organizers. Retrieved 2026-10-27.

Configuration: Qwen 3.6 - 27b

Notes: WMT26 General MT findings (2026.wmt-1.48), Figure 3: Avg. Rank 35.1 for the system listed as 'Qwen 3.6 - 27b', averaged over the 23 language pairs it entered; rank 1 is best, so lower is better (not a 0-100 score). Per-pair cESA scores (0-100, higher is better, not comparable across pairs): cs-vi 63.4, cs-de 58.3, en-cs 61.9, en-de 51.1, en-is 33.0, en-zh-Hant 59.4, en-arz 65.4, en-ru 54.4, en-be 27.6, en-lld 3.0, en-se 13.9, en-et 39.3, en-ja 68.7, en-kk 43.4, en-id 59.8, zh-ja 55.7, en-ko 55.8, en-th 42.1, en-lij 5.7, cs-uk 59.0, en-uk 44.3, en-hy 21.6, en-zh-Hans 63.8. Source: WMT26 organizers. Retrieved 2026-10-27.

Configuration: GPT OSS 20B

Notes: WMT26 General MT findings (2026.wmt-1.48), Figure 3: Avg. Rank 39.8 for the system listed as 'GPT OSS 20B', averaged over the 23 language pairs it entered; rank 1 is best, so lower is better (not a 0-100 score). Per-pair cESA scores (0-100, higher is better, not comparable across pairs): cs-vi 60.6, cs-de 45.4, en-cs 37.7, en-de 37.5, en-is 24.1, en-zh-Hant 53.3, en-arz 69.9, en-ru 49.5, en-be 21.6, en-lld 1.1, en-se 0.8, en-et 23.3, en-ja 72.4, en-kk 40.8, en-id 56.0, zh-ja 55.1, en-ko 52.0, en-th 31.3, en-lij 2.4, cs-uk 49.7, en-uk 44.6, en-hy 29.0, en-zh-Hans 51.7. Source: WMT26 organizers. Retrieved 2026-10-27.

Configuration: Qwen 3.5 - 9b

Notes: WMT26 General MT findings (2026.wmt-1.48), Figure 3: Avg. Rank 41.5 for the system listed as 'Qwen 3.5 - 9b', averaged over the 23 language pairs it entered; rank 1 is best, so lower is better (not a 0-100 score). Per-pair cESA scores (0-100, higher is better, not comparable across pairs): cs-vi 66.3, cs-de 53.7, en-cs 32.4, en-de 30.9, en-is 17.4, en-zh-Hant 51.3, en-arz 54.3, en-ru 44.4, en-be 16.5, en-lld 1.1, en-se 0.6, en-et 24.9, en-ja 57.4, en-kk 30.4, en-id 54.1, zh-ja 53.1, en-ko 18.9, en-th 37.4, en-lij 1.4, cs-uk 59.6, en-uk 48.3, en-hy 18.8, en-zh-Hans 55.4. Source: WMT26 organizers. Retrieved 2026-10-27.