MergekitCloud/mergekit-121
TEXT GENERATIONPricing:Input $0.431 / Cached $0.0862 / Output $1.12Concurrent Unit Cost:1Model Size:14.8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Oct 8, 2026Architecture:Transformer Featherless Exclusive Cold
MergekitCloud/mergekit-121 is a 14.8 billion parameter language model created using the TIES merge method, based on Qwen/Qwen2.5-14B. This model integrates Qwen/Qwen2.5-Coder-14B-Instruct, suggesting a specialization in code-related tasks. It is designed for applications requiring a robust language model with enhanced coding capabilities.
Loading preview...
Overview
MergekitCloud/mergekit-121 is a 14.8 billion parameter language model developed by MergekitCloud. It was constructed using the TIES (Trimmed, Iterative, and Selective) merge method, which combines pre-trained language models to create a new, specialized model. The base model for this merge was Qwen/Qwen2.5-14B, a powerful foundation for language understanding and generation.
Key Capabilities
- Code-centric Specialization: The primary component merged into this model is Qwen/Qwen2.5-Coder-14B-Instruct. This indicates a strong focus and optimization for code generation, understanding, and instruction-following in programming contexts.
- Mergekit Integration: Built with mergekit, this model benefits from a structured and configurable approach to model merging, allowing for precise control over the integration of different model strengths.
- Qwen2.5 Architecture: Inherits the robust architecture and capabilities of the Qwen2.5 family, known for strong performance across various language tasks.
Good For
- Code Generation: Ideal for tasks such as writing new code snippets, completing code, or translating between programming languages.
- Code Instruction Following: Excels at interpreting and executing programming-related instructions.
- Developer Tools: Suitable for integration into IDEs, code assistants, or automated programming environments.