TeichAI/gemma-4-26B-A4B-it-Claude-Opus-Distill-v2

VISIONConcurrency Cost:2Model Size:26BQuant:FP8Ctx Length:32kTool Calling:SupportedPublished:Apr 8, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

TeichAI/gemma-4-26B-A4B-it-Claude-Opus-Distill-v2 is a 26 billion parameter instruction-tuned language model built on the Gemma 4 architecture. Fine-tuned using Unsloth, it distills high-effort reasoning capabilities from Claude-4.6 Opus interactions, leveraging datasets focused on complex problem-solving. This model excels in demanding domains such as coding, scientific reasoning, deep research, and general-purpose instruction following requiring high logical coherence, with a context length of 32768 tokens.

Loading preview...

Model Overview

TeichAI/gemma-4-26B-A4B-it-Claude-Opus-Distill-v2 is a 26 billion parameter instruction-tuned model based on the Gemma 4 architecture, specifically unsloth/gemma-4-26B-A4B-it. Its primary distinction lies in its fine-tuning process, which distills advanced reasoning capabilities from Claude-4.6 Opus interactions. This was achieved by utilizing datasets where the reasoning effort was explicitly set to "High," enabling the model to break down complex problems and deliver precise solutions.

Key Capabilities

  • High-Effort Reasoning: Optimized to absorb and apply sophisticated reasoning logic, making it adept at complex problem-solving.
  • Efficient Fine-tuning: Developed using Unsloth for memory and compute optimization during the supervised fine-tuning (SFT) process.
  • Structured Thinking: Trained extensively on Claude Opus 4.6 reasoning trajectories to adopt a structured and efficient thinking pattern.
  • Multimodal Support (Base Model): While this specific fine-tune focuses on text, the underlying Gemma 4 architecture supports multimodal inputs (audio, images, video) with appropriate model variants and processing.

Ideal Use Cases

This model is particularly well-suited for applications requiring strong analytical and logical capabilities:

  • Coding: Advanced code generation, debugging, and software architecture planning.
  • Science: Deep scientific reasoning, hypothesis evaluation, and analytical problem-solving.
  • Deep Research: Navigating complex, multi-step research queries and synthesizing vast amounts of information.
  • General Purpose: Highly capable instruction-following for everyday tasks demanding high logical coherence.