IT之家 · 智能时代 · 10/8/2026, 13:54:56
PFN Launches PLaMo 3 Translate 31B: Outperforms GPT-6.1 in Japanese at Ultra-Low Cost
Preferred Networks (PFN) released the fully self-developed PLaMo 3 Translate 31B model, supporting 53 languages and online meeting translation. The model outperformed OpenAI's GPT-6.1 Sol in four benchmarks and reduced inference costs to just 14 JPY per 100k characters through efficient Japanese processing.
SOURCE COVERAGEOriginal coverage
On October 8, Japanese company Preferred Networks (PFN) announced the release of PLaMo 3 Translate 31B, a translation model optimized for Japanese while also supporting other languages. The model is entirely developed in-house by Japan, with Japanese accounting for approximately 30% of its training data.
Built upon the PLaMo 3 base model, PLaMo 3 Translate 31B expands language support from two languages in the previous generation (Japanese and English) to 53 languages. It also introduces an online meeting translation mode covering 15 languages.
PFN claims that PLaMo 3 Translate 31B achieves higher average scores than OpenAI’s GPT-6.1 Sol and GPT-6 Astra across four translation benchmarks. Thanks to its efficient handling of Japanese, the model’s cost is only ¥14 per 100,000 input characters (Note: IT Home reports this equates to approximately ¥0.59 RMB at current exchange rates), significantly lower than competing models.