# GLM-4.1V-9B-Thinking

> Z.ai's 9B open-weight vision-language reasoner that punches at 72B scale

- Maker: Z.ai (Zhipu / GLM)
- Line: GLM-V (vision-language)
- Canonical page: https://ai-tldr.dev/models/glm-4-1v-thinking/
- Updated: 2026-06-18

## In short

GLM-4.1V-9B-Thinking is an open-weight vision-language model released on July 1, 2025 by Z.ai (the Zhipu AI / GLM team) together with Tsinghua University's KEG lab. It is the first reasoning-focused entry in the GLM-V multimodal line, built on the GLM-4-9B-0414 base and paired with an AIMv2-Huge vision encoder. It has a 64K-token context window and costs $0.035 per million input tokens and $0.14 per million output tokens.

## Facts

| Fact | Value | Source |
| --- | --- | --- |
| Context window | 64K (max output 8K) | [1] |
| Input price | $0.035 / 1M tokens | [2] |
| Output price | $0.14 / 1M tokens | [2] |
| Licence | MIT · open weights | [1] |
| Released | 1 July 2025 | [3] |
| Parameters | 9B | [1] |
| Knowledge cutoff | Not disclosed | [1] |

## Sources

1. [Hugging Face model card](https://huggingface.co/zai-org/GLM-4.1V-9B-Thinking)
2. [SiliconFlow pricing & API](https://www.siliconflow.com/models/glm-4-1v-9b-thinking)
3. [GLM-V GitHub repository](https://github.com/zai-org/GLM-V)

## Overview

GLM-4.1V-9B-Thinking is an open-weight vision-language model released on July 1, 2025 by Z.ai (the Zhipu AI / GLM team) together with Tsinghua University's KEG lab. It is the first reasoning-focused entry in the GLM-V multimodal line, built on the GLM-4-9B-0414 base and paired with an AIMv2-Huge vision encoder. Despite being a roughly 9-billion-parameter model, it is positioned to compete with much larger systems on multimodal reasoning.

The defining feature is its 'thinking' paradigm: GLM-4.1V-9B-Thinking produces an explicit chain-of-thought before answering, which the team trained with a method they call Reinforcement Learning with Curriculum Sampling (RLCS). The model accepts text, images, and video, supports a 64K-token context, and handles arbitrary aspect ratios and image resolutions up to 4K via 2D-RoPE positional encoding.

Because the weights ship under an MIT license, GLM-4.1V-9B-Thinking can be downloaded, fine-tuned, and self-hosted without API fees, and its small size makes it practical to run on a single modern GPU. It is also offered as a hosted API through Z.ai's platform and third-party inference providers such as SiliconFlow for teams that prefer not to manage their own infrastructure.

## Benchmarks

- MMStar: 72.9% ([source](https://arxiv.org/abs/2507.01006))
- MathVista: 80.7% ([source](https://arxiv.org/abs/2507.01006))
- MMMU: 68% ([source](https://arxiv.org/abs/2507.01006))
- MMMU-Pro: 57.1% ([source](https://arxiv.org/abs/2507.01006))
- AI2D: 87.9% ([source](https://arxiv.org/abs/2507.01006))
- MMBench-V1.1-EN: 85.8% ([source](https://arxiv.org/abs/2507.01006))
- OCRBench: 84.2% ([source](https://arxiv.org/abs/2507.01006))
- VideoMME (with subtitles): 73.6% ([source](https://arxiv.org/abs/2507.01006))
- WeMath: 63.8% ([source](https://arxiv.org/abs/2507.01006))

## Pricing

- Input: $0.035 / 1M tokens per 1M tokens
- Output: $0.14 / 1M tokens per 1M tokens

Hosted API pricing via SiliconFlow. The model weights are open under MIT, so self-hosting incurs no per-token fee.

Source: https://www.siliconflow.com/models/glm-4-1v-9b-thinking

## Strengths

- Open-weight (MIT) — free to download, fine-tune, and self-host with no usage restrictions
- Strong multimodal reasoning for its size: leads 10B-class models on 23 of 28 reported benchmarks and beats Qwen2.5-VL-72B on 18 of them
- Explicit chain-of-thought 'thinking' output improves accuracy and interpretability on STEM and math-vision tasks
- Handles arbitrary aspect ratios and up to 4K image resolution, plus video input
- Small enough (9B) to run on a single GPU, keeping inference cheap

## Best for

- Solving STEM and math problems presented as images, diagrams, or charts
- Document and chart understanding, including OCR-heavy and long-document pages
- Video understanding and question answering
- GUI / screenshot understanding and UI navigation agents
- Self-hosted multimodal reasoning where data privacy or cost rules out a closed API

## How to access

- [Z.ai (Zhipu / BigModel)](https://docs.z.ai/guides/vlm/glm-4.1v-thinking) — model id `glm-4.1v-thinking-flash`
- [SiliconFlow](https://www.siliconflow.com/models/glm-4-1v-9b-thinking) — model id `zai-org/GLM-4.1V-9B-Thinking`
- [Hugging Face (weights)](https://huggingface.co/zai-org/GLM-4.1V-9B-Thinking) — model id `zai-org/GLM-4.1V-9B-Thinking`

## GLM-V (vision-language) versions

- [GLM-4.6V](https://ai-tldr.dev/models/glm-4-6v/) (8 December 2025)
- [GLM-4.5V](https://ai-tldr.dev/models/glm-4-5v/) (11 August 2025)
- GLM-4.1V-9B-Thinking (1 July 2025) — this page

## FAQ

### Is GLM-4.1V-9B-Thinking open source?

Yes. The weights are released under the MIT license on Hugging Face (zai-org/GLM-4.1V-9B-Thinking), so you can download, fine-tune, and self-host the model commercially without per-token fees.

### What can GLM-4.1V-9B-Thinking process besides text?

It is a vision-language model that accepts images and video alongside text, with text output. It handles arbitrary aspect ratios and image resolutions up to 4K, and supports a 64K-token context window.

### How does a 9B model compete with 72B models?

Its 'thinking' chain-of-thought paradigm, trained with Reinforcement Learning with Curriculum Sampling (RLCS), lets it reason step by step before answering. In Z.ai's technical report it leads 10B-class models on 23 of 28 benchmarks and outperforms the much larger Qwen2.5-VL-72B on 18 of them.

### How much does GLM-4.1V-9B-Thinking cost to use?

Self-hosting is free since the weights are open. As a hosted API it is inexpensive — SiliconFlow lists about $0.035 per million input tokens and $0.14 per million output tokens, and it is also served through Z.ai's own platform.

## Links

- [Hugging Face model card](https://huggingface.co/zai-org/GLM-4.1V-9B-Thinking)
- [GLM-V GitHub repository](https://github.com/zai-org/GLM-V)
- [Technical report (arXiv 2507.01006)](https://arxiv.org/abs/2507.01006)
- [Z.ai vision API docs](https://docs.z.ai/guides/vlm/glm-4.1v-thinking)
- [SiliconFlow pricing & API](https://www.siliconflow.com/models/glm-4-1v-9b-thinking)
- [Live demo (Hugging Face Space)](https://huggingface.co/spaces/zai-org/GLM-4.1V-9B-Thinking-Demo)

Source: AI/TLDR LLM registry, https://ai-tldr.dev/models/glm-4-1v-thinking/