KinyarwandaSpeech
Text-to-Speech (TTS)
Currently the highest-quality synthetic voice for Kinyarwanda, enabling natural-sounding interactions for voice-based applications.
Models
Datasets
Models
Open-Source AI Models
A suite of high-quality, open-source models developed for African languages, retrieval workflows, speech, and general NLP use cases.
Model releases
11
Primary focus
Speech + NLP
Language coverage
Kinyarwanda, Swahili, Yoruba, Hausa
RetrievalRAG
KinyaCOLBERT Free
The first Kinyarwanda embedding and information retrieval model. Essential for building RAG-based chatbots like Tunga.
KinyarwandaNLP
KinyaBERT
A foundational BERT-style model serving as the backbone for various Kinyarwanda NLP tasks, from classification to entity recognition.
Swahili8B
UlizaLlama3
An 8B parameter model by Jacaranda specifically enhanced to excel in processing and generating text in Swahili.
Yoruba8B
YorubaLlama
An 8B parameter model by Jacaranda specifically enhanced to excel in processing and generating text in the Yoruba language.
Hausa8B
HausaLlama
An 8B parameter model by Jacaranda specifically enhanced to excel in processing and generating text in the Hausa language.
isiXhosaisiZulu
Xhosa_ZuluLlama3
An 8B parameter model by Jacaranda specifically enhanced to excel in processing and generating text in isiXhosa and isiZulu.
Multilingual8B
AfroLlama_V1
An 8B parameter multi-language model by Jacaranda enhanced for Swahili, Xhosa, Zulu, Yoruba, and Hausa.
Swahili7B
UlizaLlama
A 7B parameter Swahili language model by Jacaranda continually pre-trained on Swahili instructions.
SwahiliFoundation
Kiswallama Pretrained
A Swahili foundational model continually pre-trained to extend the capabilities of Llama 2.
CommunityMultilingual
Masakhane Models
Browse open-source models from the Masakhane community spanning multiple African languages and NLP tasks.