User Claims Qwen3-VL-2B is the Only Viable VLM for JSON Extraction on Low-End Hardware
English summary
A Reddit user reports that after extensive testing on three low-end laptops (Intel i3, 8GB RAM, integrated GPU), Qwen3-VL-2B in Q4_K_M GGUF quantization reliably extracts data from images to JSON, outperforming Qwen3-VL-4B and Qwen3.5 2B. The user notes this model is absent from major benchmarks like Artificial Analysis and the Open LLM Leaderboard, which list the 4B version instead. The post questions why it is ignored and asks if any other model can handle the task on similarly constrained devices like phones or Raspberry Pis. No quantitative benchmarks or replication details are provided.
Chinese summary
一位Reddit用户报告，在三台低端笔记本电脑（Intel i3、8GB内存、集成显卡）上大量测试后，Qwen3-VL-2B的Q4_K_M GGUF量化版本能可靠地从图像中提取数据生成JSON，表现优于Qwen3-VL-4B和Qwen3.5 2B。该用户指出此模型未出现在Artificial Analysis或Open LLM Leaderboard等主流基准测试中（它们只列出4B版本），并质疑为何被忽视，询问是否有其他模型能在手机或树莓派等类似受限设备上完成该任务。未提供量化基准或复现细节。
Key points
A user found Qwen3-VL-2B Q4_K_M GGUF reliably extracts JSON from images on low-end laptops (i3, 8GB RAM).
用户发现Qwen3-VL-2B Q4_K_M GGUF能在低端笔记本电脑（i3、8GB内存）上可靠地从图像提取JSON。
In informal testing, it outperformed Qwen3-VL-4B and Qwen3.5 2B for this specific data extraction task.
在非正式测试中，它在这一特定数据提取任务上优于Qwen3-VL-4B和Qwen3.5 2B。
The model is missing from major leaderboards (Artificial Analysis, Open LLM Leaderboard) while the 4B version is listed.
该模型未出现在主要排行榜（Artificial Analysis、Open LLM Leaderboard）中，而4B版本被列出。
The user asks why it is ignored and whether alternatives exist for phones or Raspberry Pis.
用户质疑为何被忽视，并询问在手机或树莓派上是否有其他替代方案。