Hugging Face's logo

inclusionAI
/
ASearcher-Web-14B

PyTorch
qwen2
Model card Files
xet
Community
1
ASearcher-Web-14B
File size: 3,229 Bytes 
a8a100c
 
 
 
 
 
 
 
caf4195
 
 
 
 
 
 
 
 
 
 
a8a100c
caf4195
 
 
a1cd42c
a8a100c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2bae89b
a8a100c
 
2bae89b
a8a100c
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
 
---
license: apache-2.0
datasets:
- inclusionAI/ASearcher-train-data
base_model:
- Qwen/Qwen2.5-14B
---

<style>
  .no-border-table table, .no-border-table th, .no-border-table td {
    border: none !important;
  }
</style>

<div class="no-border-table">

| | |
|-|-|
| [![GitHub](https://img.shields.io/badge/GitHub-Repository-black?logo=github)](https://github.com/inclusionAI/ASearcher) | [![arXiv](http://img.shields.io/badge/cs.AI-arXiv%3A2506.07982-B31B1B.svg?logo=arxiv&logoColor=red)](https://arxiv.org/abs/2508.07976) |

</div>

### Instruction

ASearcher is an open-source framework designed for large-scale online reinforcement learning (RL) training of search agents. Our mission is to advance Search Intelligence to expert-level performance. We are fully committed to open-source by releasing model weights, detailed training methodologies, and data construction pipelines. Additionally, we provide comprehensive guidance on building and training customized agents based on AReaL. ASearcher empowers developers to build their own high-performance search agents easily and cost-effectively.

We have released multiple models trained with different settings and based on foundation models of varying sizes. These models have achieved outstanding performance on Single-Hop / Multi-Hop QA and more challenging tool-augmented benchmarks like GAIA, Xbench.

### Model Download

| Model Name | Base Model | Training Setting | Download Link |
|------------|----------------|------------------|----------------|
| ASearcher-Local-7B | Qwen2.5-7B | Local knowledge base with RAG | 🤗[Huggingface](https://huggingface.co/inclusionAI/ASearcher-Local-7B) |
| ASearcher-Web-7B | Qwen2.5-7B | Web-based search and browsing | 🤗[Huggingface](https://huggingface.co/inclusionAI/ASearcher-Local-14B) |
| ASearcher-Local-14B | Qwen2.5-14B | Local knowledge base with RAG | 🤗[Huggingface](https://huggingface.co/inclusionAI/ASearcher-Local-7B) |
| ASearcher-Web-14B | Qwen2.5-14B | Web-based search and browsing | 🤗[Huggingface](https://huggingface.co/inclusionAI/ASearcher-Local-7B) |
| ASearcher-Web-QwQ-32B | QwQ-32B | Web-based search and browsing | 🤗[Huggingface](https://huggingface.co/inclusionAI/ASearcher-Local-7B) |

### Performance
#### Evaluation on challenging benchmarks (ASearcher-Web-QwQ)
<img src="https://cdn-uploads.huggingface.co/production/uploads/63159678915d0b80682fe9f9/ds_PHc5cD0PXO8TmiY7_U.png" style="width:50%;">

#### Evaluation with a local knowledge base with RAG
<img src="https://cdn-uploads.huggingface.co/production/uploads/63159678915d0b80682fe9f9/SkWaECNeJUHGF4BJYMKWz.png" style="width:50%;">

#### Evaluation with web-based search and browsing
<img src="https://cdn-uploads.huggingface.co/production/uploads/63159678915d0b80682fe9f9/SmRjd6RsrTiWAY9957-1Z.png" style="width:50%;">

### Dataset Download
We also release our full [training data](https://huggingface.co/datasets/inclusionAI/ASearcher-train-data) and [test data](https://huggingface.co/datasets/inclusionAI/ASearcher-test-data), you can easily get them and reproduce our result.

### Quickstart
If you want to learn more details, please refer to our GitHub repository: [ASearcher](https://github.com/inclusionAI/ASearcher)