Multimodal CAPTCHA-Solving Agent Training Data

ReCAP-187K-SFT contains supervised fine-tuning data for training multimodal GUI agents to solve CAPTCHAs. The dataset is structured in Qwen3-style conversation format and includes references to screenshot images from interaction trajectories. It was created by ReCAP-Agent and last updated in March 2026.

Use Cases

• Fine-tune multimodal language models on Qwen3-style conversation data for CAPTCHA-solving agent tasks.
• Train GUI agents using referenced screenshot images to learn visual interaction trajectories.
• Develop and benchmark models for multimodal tasks combining text conversation data with image-based references.

Strengths

• Dataset is specifically designed for the focused task of CAPTCHA-solving agent training.
• Data is formatted in a structured Qwen3-style conversation format suitable for model fine-tuning.
• Includes multimodal elements with text conversations and references to screenshot images.

Limitations

• Unknown total row count and column structure limits precise assessment of dataset scale and features.
• Sample data and file formats are unavailable, making initial inspection and integration planning difficult.
• Potential for task-specific bias, as the data is exclusively focused on CAPTCHA-solving interactions.

Provenance

Source

ReCAP-Agent on Hugging Face.

Collection Method

null

Time Range

null

Freshness

Last updated March 2026.

Geography

Tags suggest a US region focus.