---

license: apache-2.0

base_model:

• apus-ailab/APUS-OpenJev-v1-9B

tags:

• text-generation-inference
• llama-cpp
• apus-openjev
• decision-model
• structured-output
• bf16

language:

• en
• zh

pipeline_tag: text-generation

library_name: transformers

---

APUS-OpenJev-v1-9B-GGUF

> APUS-OpenJev-v1-9B is a Qwen3.5-9B-based decision model from APUS AI-LAB, designed for browser action selection, workflow routing, and natural-language principle judgments rather than open-ended text generation — this repository provides the 9B checkpoint-3000 merged BF16 weights, downloadable independently without a separate LoRA adapter. It scores dynamic candidates supplied per request and returns their preference distribution, reusing Qwen's language representations and vocabulary projection so application code can assemble decisions into structured workflow outputs; its native runtime supports two effort levels — low (16 layers) and high (32 layers, recommended for text generation). On the internal Frozen80 development panel (covering Browser, HelpSteer3, BoolQ, MNLI, and attribute-decision tasks), the merged model scores 68/80 (85.00%), though the authors note this is a reused engineering panel rather than an independent blind benchmark or end-to-end browser success rate, and that candidate probabilities express relative preference rather than calibrated correctness (with BF16 merging further shifting some probabilities). It's part of a three-size family (4B, 9B, 35B-A3B) hosted in separate repositories under a shared collection, with the 4B release instead selecting checkpoint-5949 (82.50% on the same panel), and is released under Apache 2.0 as a finetune of the Qwen3.5-9B base.

Model Files

| File Name | Quant Type | File Size | File Link | Description |

|-----------|------------|-----------|-----------|-------------|

| APUS-OpenJev-v1-9B.BF16.gguf | BF16 | 17.9 GB | Link | Full BF16 weights. Highest quality, largest file size. |

| APUS-OpenJev-v1-9B.Q3_K_L.gguf | Q3_K_L | 4.93 GB | Link | Lower quality but usable, good for low RAM availability. |

| APUS-OpenJev-v1-9B.Q3_K_M.gguf | Q3_K_M | 4.62 GB | Link | Low quality. |

| APUS-OpenJev-v1-9B.Q4_K_M.gguf | Q4_K_M | 5.63 GB | Link | Good quality, default size for most use cases, recommended. |

| APUS-OpenJev-v1-9B.Q4_K_S.gguf | Q4_K_S | 5.35 GB | Link | Slightly lower quality with more space savings, recommended. |

| APUS-OpenJev-v1-9B.Q5_K_M.gguf | Q5_K_M | 6.47 GB | Link | High quality, recommended. |

| APUS-OpenJev-v1-9B.Q5_K_S.gguf | Q5_K_S | 6.31 GB | Link | High quality, recommended. |

| APUS-OpenJev-v1-9B.Q6_K.gguf | Q6_K | 7.36 GB | Link | Very high quality, near perfect, recommended. |

llama.cpp

LLM inference in C/C++ — https://github.com/ggml-org/llama.cpp