Fox-AI-by-teolm30/fox1.4
TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:0.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Apr 4, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold
Fox1.4 by teolm30 is a 900 million parameter language model, built upon the Qwen2.5-0.5B base model and merged with a LoRA adapter. It is specifically trained on combined data from math, logic, knowledge, and code reasoning tasks, making it a specialist in these domains. This model excels at complex reasoning challenges, demonstrating 100% accuracy on a custom 10-question benchmark covering various reasoning tasks.
Loading preview...
Fox1.4: A Reasoning Specialist
Fox1.4, developed by teolm30, is a 900 million parameter language model designed for advanced reasoning tasks. It is built on the Qwen2.5-0.5B base model, enhanced with a LoRA adapter, and specifically trained on a diverse dataset encompassing math, logic, general knowledge, and code reasoning.
Key Capabilities
- Specialized Reasoning: Optimized for complex problem-solving across multiple domains.
- High Accuracy on Reasoning Tasks: Achieved 100% on a custom 10-question benchmark, successfully handling tasks like penguin exception logic, riddles, arithmetic, factual knowledge, and basic code functions.
- Estimated MMLU Score: Approximately 40-50%, indicating a solid understanding of various subjects.
- Compact and Efficient: With 900 million parameters, it offers strong reasoning capabilities in a relatively smaller footprint.
Good For
- Applications requiring robust mathematical and logical problem-solving.
- Tasks involving knowledge-based question answering.
- Scenarios needing code reasoning and understanding.
- Developers looking for a specialized model for analytical and deductive tasks.