VTC-R1: Vision-Text Compression for Efficient Long-Context Reasoning

TL;DR

VTC-R1 introduces a vision-text compression method that converts reasoning segments into images, enabling more efficient long-context reasoning in vision-language models with significant speed improvements.

Contribution

It proposes a novel vision-text compression paradigm that enhances reasoning efficiency by rendering reasoning steps into images for iterative processing.

Findings

Achieves 3.4x token compression on a new dataset.

Outperforms standard long-context reasoning benchmarks.

Provides 2.7x speedup in inference latency.

Abstract

Long-context reasoning has significantly empowered large language models (LLMs) to tackle complex tasks, yet it introduces severe efficiency bottlenecks due to the computational complexity. Existing efficient approaches often rely on complex additional training or external models for compression, which limits scalability and discards critical fine-grained information. In this paper, we propose VTC-R1, a new efficient reasoning paradigm that integrates vision-text compression into the reasoning process. Instead of processing lengthy textual traces, VTC-R1 renders intermediate reasoning segments into compact images, which are iteratively fed back into vision-language models as "optical memory." We construct a training dataset based on OpenR1-Math-220K achieving 3.4x token compression and fine-tune representative VLMs-Glyph and Qwen3-VL. Extensive experiments on benchmarks such as MATH500, AIME25, AMC23 and GPQA-D demonstrate that VTC-R1 consistently outperforms standard long-context reasoning. Furthermore, our approach significantly improves inference efficiency, achieving 2.7x speedup in end-to-end latency, highlighting its potential as a scalable solution for reasoning-intensive applications. Our code is available at https://github.com/w-yibo/VTC-R1.