SL-AI/GRaPE-2-Pro

Hugging Face
VISIONConcurrent Unit Cost:2Model Size:27BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Apr 19, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Warm

SL-AI/GRaPE-2-Pro is a 27 billion parameter multimodal language model developed by Skinnertopia Lab for Artificial Intelligence (SLAI), built on the Qwen3.5 architecture. It accepts image and text inputs to produce text outputs, featuring an extended thinking mode system for controllable reasoning depth. Optimized for code, STEAM, logical reasoning, and structured problem solving, GRaPE-2-Pro is designed for large-scale intelligence and raw reasoning tasks.

Loading preview...

GRaPE 2 Pro: Flagship Multimodal Reasoning Model

GRaPE 2 Pro is the flagship 27 billion parameter model from SLAI's second-generation GRaPE family, built upon the robust Qwen3.5 architecture. This multimodal model processes both image and text inputs to generate text outputs, distinguishing itself with an innovative extended thinking mode system that allows for controllable reasoning depth.

Key Capabilities & Differentiators

  • Stronger Base Model: Leverages Qwen3.5-27 for enhanced performance.
  • Expanded Thinking Modes: Features six discrete reasoning tiers (minimal, low, medium, high, xtra-Hi, auto) controllable via the <thinking_mode> tag, enabling users to specify the depth of reasoning for complex tasks.
  • Proprietary Training Data: Post-trained on a curated, closed-source dataset with significant emphasis on:
    • Code (~50% of post-training data)
    • STEAM (Science, Technology, Engineering, Arts, and Mathematics)
    • Logical reasoning and structured problem solving
  • Optimized for Reasoning: Designed to excel in structured reasoning tasks, making it suitable for analytical and problem-solving applications.

When to Use GRaPE 2 Pro

GRaPE 2 Pro is ideal for use cases requiring:

  • Complex Code Generation: Utilize high or xtra-Hi thinking modes for intricate coding challenges.
  • Multi-step Mathematical Problems: Benefit from its enhanced logical reasoning capabilities.
  • Deep Analytical Work: Engage extended reasoning for thorough analysis.
  • Agentic Applications: Low or Auto thinking modes are recommended to balance reasoning with action speed.

Finetunes
2 models