McGill-NLP/AfriqueQwen3.5-4B-50Langs-Instruct-v1

VISIONPricing:Input $0.4 / Cached $0.08 / Output $0.8Concurrent Unit Cost:1Model Size:4.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 30, 2026License:cc-by-4.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

The McGill-NLP/AfriqueQwen3.5-4B-50Langs-Instruct-v1 is a 4.5 billion parameter instruction-tuned language model developed by McGill-NLP, based on the Qwen3.5 architecture with a 32768 token context length. This model is specifically optimized for performance across 50 African languages, demonstrating strong capabilities in multilingual tasks including mathematics, translation, and multiple-choice questions. It significantly outperforms other models in its class on various African language benchmarks such as AfriMGSM, AfriMMLU, and SIB-200.

Loading preview...

What is AfriqueQwen3.5-4B-50Langs-Instruct-v1?

McGill-NLP/AfriqueQwen3.5-4B-50Langs-Instruct-v1 is a 4.5 billion parameter instruction-tuned language model developed by McGill-NLP. It is a post-trained version of the AfriqueQwen3.5-4B-50Langs base model, designed to excel in a wide array of tasks across 50 African languages. With a substantial context length of 32768 tokens, it is well-suited for processing longer inputs and generating comprehensive responses.

Key Capabilities & Performance

This model demonstrates superior performance on various African language benchmarks compared to other models in its size class. Key highlights include:

  • Multilingual Proficiency: Achieves an overall score of 61.6 on a comprehensive suite of African language benchmarks, including AfriMGSM, AfriMMLU, AfriXNLI, Belebele, Injongo, SIB-200, and FLORES.
  • Mathematical Reasoning: Scores 54.2 on AfriMGSM, indicating strong capabilities in solving math word problems in African languages.
  • Language Understanding: Records 45.4 on AfriMMLU and 62.2 on AfriXNLI, showcasing robust multilingual understanding.
  • Instruction Following: Optimized for instruction-following tasks, providing step-by-step reasoning for complex problems.

When to Use This Model

This model is particularly well-suited for applications requiring strong performance in African languages. Consider using it for:

  • Multilingual Chatbots: Developing conversational AI agents that can interact effectively in diverse African linguistic contexts.
  • Educational Tools: Creating resources for mathematics and general knowledge in African languages.
  • Content Generation & Translation: Tasks involving text generation, summarization, or translation across the 50 supported languages.
  • Research & Development: As a robust foundation for further fine-tuning on specific African language datasets or tasks.