Tether AI Research unveiled offline translation models for phones, including AfriSLM for 19 African languages and EuroNano for nine European languages. Both run translations on-device, keeping personal data local and working without a network connection.
AfriSLM supports Hausa, Amharic, Yoruba, Lingala, Swahili, Igbo, Zulu, Somali, Oromo, Malagasy, Kinyarwanda, Xhosa, Afrikaans, Wolof, Luganda, Nyanja, Shona, Tswana and Southern Sotho. Tether estimates the languages have a combined reach of about 50% of the continent’s population; this is language reach, not a count of individual users. Tether says the 800-million-parameter model outperformed Qwen3.5-122B-A10B, TranslateGemma-27B and NLLB-3.3B in evaluations using FLORES-200, BOUQuET and SMOL; benchmark results do not guarantee the same performance in every translation situation. The company says it discarded up to 96% of training material judged unsuitable and offers 0.8B, 2B and 4B versions.
EuroNano uses English as an intermediary to connect nine European languages across 90 translation directions and takes up 36MB. Tether says that is about 94% less storage than a comparable Firefox offline arrangement. AfriSLM is available through Hugging Face, and its associated research has been accepted for presentation at EMNLP 2026.
