Qwen3-4B-Thinking-2507-DAG-GGUF

The sequelbox/Qwen3-4B-Thinking-2507-DAG-Reasoning model is an experimental specialist reasoning AI finetuned for multi-step causal analysis and reasoning, producing structured Directed Acyclic Graphs (DAGs) in response to user input across fields like programming, science, business, law, and more; it structures its output in a readable JSON graph format with confidence measures, enabling easy visualization and further analysis, and supports prompt-driven graph generation using a custom Qwen3-4B-Thinking-2507 inference approach, making it suitable for advanced reasoning tasks on both desktop and server environments; the model is part of open-source research, benefits from a custom DAG dataset, and is recommended for those needing clear causal structure in analysis or automation outputs.

Model Files

File Name	Quant Type	File Size
Qwen3-4B-Thinking-2507-DAG-Reasoning.BF16.gguf	BF16	8.05 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.F16.gguf	F16	8.05 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.F32.gguf	F32	16.1 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.Q2_K.gguf	Q2_K	1.67 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.Q3_K_L.gguf	Q3_K_L	2.24 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.Q3_K_M.gguf	Q3_K_M	2.08 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.Q3_K_S.gguf	Q3_K_S	1.89 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.Q4_K_M.gguf	Q4_K_M	2.5 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.Q4_K_S.gguf	Q4_K_S	2.38 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.Q5_K_M.gguf	Q5_K_M	2.89 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.Q5_K_S.gguf	Q5_K_S	2.82 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.Q6_K.gguf	Q6_K	3.31 GB
Qwen3-4B-Thinking-2507-DAG-Reasoning.Q8_0.gguf	Q8_0	4.28 GB

Quants Usage

(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)

Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better):