nlpguy/Qwen3.8-27B-Fimi-2

VISIONPricing:Input $1.6 / Cached $0.15 / Output $12Concurrent Unit Cost:2Model Size:27BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 23, 2026Architecture:Transformer0.0K Featherless Exclusive Cold

nlpguy/Qwen3.8-27B-Fimi-2 is a 27 billion parameter language model based on the Qwen3.8 architecture, created by nlpguy through a merge of several pre-trained models. Utilizing the Reinforced Agent Merging Plus (Tensor-Local) method, this model integrates capabilities from DavidAU/Qwen3.8-27B-Cold-Fusion-GAIN-V1.1, TeichAI/Qwen3.8-27B-Fable-Distill, and beyoru/Kiwen1.1-27B, with Qwen/Qwen3.8-27B as its base. It is designed to leverage the combined strengths of its constituent models, offering a versatile foundation for various natural language processing tasks.

Loading preview...

Overview

Fimi 2 is a 27 billion parameter language model developed by nlpguy, built upon the Qwen3.8 architecture. This model was created using the Reinforced Agent Merging Plus (Tensor-Local) method, a technique designed to combine the strengths of multiple pre-trained models into a single, more capable entity. The merging process utilized Qwen/Qwen3.8-27B as the foundational base model.

Merge Details

Fimi 2 integrates components from three distinct Qwen3.8-27B variants:

The merge was configured to combine specific layers from these models, including language model layers and visual blocks, aiming to synthesize their respective capabilities. This approach allows Fimi 2 to potentially inherit diverse strengths from its constituent models, making it a versatile option for general-purpose language tasks.

Intended Use

As a merged model, Fimi 2 is suitable for a broad range of applications where a robust 27B parameter model is beneficial. Its construction from multiple specialized models suggests a balanced performance across various NLP benchmarks, making it a strong candidate for tasks requiring comprehensive language understanding and generation.