Total runs:
325.8K
Run Growth:
245.1K
Growth Rate:
76.97%
Sponsored by
Agentkit
AtomicChat / qwen-agentworld-35b-GGUF
Model's Last Updated: July 23 2026
text-generation
Introduction of qwen-agentworld-35b-GGUF
Model Details of qwen-agentworld-35b-GGUF
Qwen AgentWorld 35B A3B , self-quantized to GGUF by Atomic Chat . Built straight from Qwen's original weights with a per-tensor importance matrix. Runs fully offline.
Highlights
- A language world model for agentic environment simulation from Qwen, trained across seven interaction domains: MCP, Search, Terminal, SWE, Android, Web and OS.
-
Sparse Mixture-of-Experts
: 256 routed experts with 8 active per token plus a shared expert, across 40 layers (
qwen3_5_moe). -
35B total parameters with only 3B activated per token
, built on
Qwen3.5-35B-A3B-Base. - Hybrid architecture using Gated DeltaNet and Gated Attention components.
- 262,144-token native context for long-horizon agent traces.
- Pure open : Apache 2.0 license.
-
Full quant ladder
with an importance matrix on every quant over
calibration_datav3.
These GGUFs are self-quantized from the original weights , not a repack. The importance matrix keeps low-bit quants closer to the full-precision model.
Always pass
--jinjaso the Qwen AgentWorld 35B A3B chat template is applied. Without it the model can emit malformed turns.
Model Overview
| Property | Value |
|---|---|
| Base model |
Qwen/Qwen-AgentWorld-35B-A3B
|
| Total / active parameters | 35B total / 3B activated per token |
| Layers | 40 |
| Experts | 256 routed + 1 shared, 8 active per token |
| Context length | 262,144 |
| Architecture |
qwen3_5_moe
sparse MoE (Gated DeltaNet + Gated Attention)
|
| This repo | GGUF quants (imatrix), full ladder from the original weights |
Scores are Qwen's published results for the base
Qwen/Qwen-AgentWorld-35B-A3B
. These are full-precision scores from Qwen's AgentWorldBench; the quants here run the same model locally. Quantization preserves the large majority of this, with
Q4_K_M
and up sitting within a point or two of full precision.
Choosing a quant
| Quant | Size | Notes |
|---|---|---|
Q4_K_M
|
21.2 GB | Recommended default. Best balance of size, speed and quality. |
UD-Q4_K_XL
|
21.5 GB | Dynamic. Token embeddings and output kept at Q8_0 for higher quality at a Q4 footprint. |
Q5_K_M
|
24.7 GB | Higher quality, low loss. |
Q6_K
|
28.5 GB | Near lossless. |
Q8_0
|
36.9 GB | Effectively lossless, reference quality. |
Pick the largest file that fits your (V)RAM with room for context.
Q4_K_MorUD-Q4_K_XLis the sweet spot for most setups;Q6_KorQ8_0for maximum fidelity. As an MoE the routed experts dominate the file size, so the quants are smaller than a dense 35B.
Get started
Run Qwen AgentWorld 35B A3B locally with:
-
Atomic Chat
:
the easiest path. Open the app, search
AtomicChat/qwen-agentworld-35b-GGUF, pick a quant, hit Use this model . -
llama.cpp:
llama-server -hf AtomicChat/qwen-agentworld-35b-GGUF:Q4_K_M --jinja -c 8192 -
Ollama:
ollama run hf.co/AtomicChat/qwen-agentworld-35b-GGUF:Q4_K_M - LM Studio / Jan: search the repo id, download any quant.
Best practices
| Parameter | Value |
|---|---|
| temperature | 0.6 |
| top_p | 0.95 |
| top_k | 20 |
Qwen3.5-style defaults; see the base model card for task-specific recommendations.
Run in llama.cpp
git clone https://github.com/ggerganov/llama.cpp
cmake llama.cpp -B llama.cpp/build -DBUILD_SHARED_LIBS=OFF -DGGML_CUDA=ON
cmake --build llama.cpp/build --config Release -j --target llama-cli llama-server
./llama.cpp/build/bin/llama-server \
-hf AtomicChat/qwen-agentworld-35b-GGUF:UD-Q4_K_XL \
--jinja -ngl 99 -c 8192 -fa on
How these were made
-
Download
Qwen/Qwen-AgentWorld-35B-A3B(original weights). - Convert to f16 GGUF with llama.cpp ; the NextN/MTP block is stripped before quantizing.
-
Build an importance matrix over
calibration_datav3withllama-imatrix. -
Quantize the full ladder with
--imatrix. -
UD-Q4_K_XLadditionally pins the token-embedding and output tensors toQ8_0.
License
Released by Qwen under the Apache 2.0 license. Quantized by Atomic Chat.
Runs of AtomicChat qwen-agentworld-35b-GGUF on huggingface.co
198
Total runs
-3
24-hour runs
-52
3-day runs
-176
7-day runs
-421
30-day runs
More Information About qwen-agentworld-35b-GGUF huggingface.co Model
Url of qwen-agentworld-35b-GGUF
qwen-agentworld-35b-GGUF huggingface.co Url
Provider of qwen-agentworld-35b-GGUF huggingface.co
Other API from AtomicChat
Updated:August 09 2026
Total runs:
286.5K
Run Growth:
271.8K
Growth Rate:
88.42%
Total runs:
115.0K
Run Growth:
89.2K
Growth Rate:
100.00%
Total runs:
90.1K
Run Growth:
80.4K
Growth Rate:
89.81%
Total runs:
72.9K
Run Growth:
50.0K
Growth Rate:
68.67%
Total runs:
72.7K
Run Growth:
46.7K
Growth Rate:
64.33%
Total runs:
72.4K
Run Growth:
47.8K
Growth Rate:
66.01%
Total runs:
60.1K
Run Growth:
57.1K
Growth Rate:
94.90%
Total runs:
7.5K
Run Growth:
-1.5K
Growth Rate:
-20.45%
Total runs:
5.9K
Run Growth:
740
Growth Rate:
12.52%
Total runs:
5.3K
Run Growth:
266
Growth Rate:
5.05%
Total runs:
4.6K
Run Growth:
-2.6K
Growth Rate:
-57.12%
Total runs:
4.6K
Run Growth:
-15.2K
Growth Rate:
-332.39%
Total runs:
4.5K
Run Growth:
-13.4K
Growth Rate:
-300.70%
Total runs:
4.4K
Run Growth:
-9.9K
Growth Rate:
-223.64%
Total runs:
3.6K
Run Growth:
-2.4K
Growth Rate:
-65.62%
Total runs:
2.8K
Run Growth:
453
Growth Rate:
16.58%
Total runs:
2.7K
Run Growth:
-870
Growth Rate:
-31.91%
Total runs:
2.5K
Run Growth:
-2.6K
Growth Rate:
-103.91%
Total runs:
2.5K
Run Growth:
-1.4K
Growth Rate:
-61.71%
Total runs:
2.3K
Run Growth:
-13.5K
Growth Rate:
-594.28%
Total runs:
2.0K
Run Growth:
-9.0K
Growth Rate:
-438.69%
Total runs:
2.0K
Run Growth:
325
Growth Rate:
15.97%
Total runs:
1.9K
Run Growth:
1.1K
Growth Rate:
56.75%
Total runs:
1.9K
Run Growth:
-14.8K
Growth Rate:
-787.94%
Total runs:
1.7K
Run Growth:
657
Growth Rate:
37.65%
Total runs:
1.6K
Run Growth:
703
Growth Rate:
43.08%
Total runs:
1.6K
Run Growth:
1.6K
Growth Rate:
100.00%
Total runs:
1.6K
Run Growth:
-516
Growth Rate:
-32.53%
Total runs:
1.4K
Run Growth:
1.3K
Growth Rate:
100.00%
Total runs:
1.3K
Run Growth:
-1.9K
Growth Rate:
-143.31%
Total runs:
1.2K
Run Growth:
1.2K
Growth Rate:
100.00%
Total runs:
1.0K
Run Growth:
1.0K
Growth Rate:
100.00%
Total runs:
988
Run Growth:
988
Growth Rate:
100.00%
Total runs:
982
Run Growth:
982
Growth Rate:
100.00%
Total runs:
952
Run Growth:
-351
Growth Rate:
-36.87%
Total runs:
770
Run Growth:
-198
Growth Rate:
-25.71%
Total runs:
715
Run Growth:
-1.3K
Growth Rate:
-182.24%
Total runs:
680
Run Growth:
-3.5K
Growth Rate:
-520.29%
Total runs:
620
Run Growth:
-7.6K
Growth Rate:
-1260.73%
Total runs:
602
Run Growth:
602
Growth Rate:
100.00%
Total runs:
569
Run Growth:
-81
Growth Rate:
-14.24%
Total runs:
562
Run Growth:
562
Growth Rate:
100.00%
Total runs:
557
Run Growth:
-3.3K
Growth Rate:
-636.69%
Total runs:
530
Run Growth:
14
Growth Rate:
2.64%
Total runs:
529
Run Growth:
-445
Growth Rate:
-84.12%
Total runs:
476
Run Growth:
-507
Growth Rate:
-106.51%
Total runs:
429
Run Growth:
-426
Growth Rate:
-99.30%
Total runs:
420
Run Growth:
404
Growth Rate:
96.19%
Total runs:
414
Run Growth:
-1.4K
Growth Rate:
-337.44%
Total runs:
397
Run Growth:
-414
Growth Rate:
-104.28%
Total runs:
390
Run Growth:
390
Growth Rate:
100.00%
Total runs:
378
Run Growth:
-1.4K
Growth Rate:
-371.43%
Total runs:
372
Run Growth:
-1.4K
Growth Rate:
-369.62%
Total runs:
365
Run Growth:
365
Growth Rate:
100.00%
Total runs:
363
Run Growth:
15
Growth Rate:
4.13%
Total runs:
296
Run Growth:
186
Growth Rate:
74.10%
Total runs:
291
Run Growth:
-609
Growth Rate:
-209.28%
Total runs:
283
Run Growth:
283
Growth Rate:
100.00%
Total runs:
282
Run Growth:
-1.4K
Growth Rate:
-506.03%
Total runs:
277
Run Growth:
277
Growth Rate:
100.00%
Total runs:
271
Run Growth:
271
Growth Rate:
100.00%
Total runs:
259
Run Growth:
-676
Growth Rate:
-261.00%
Total runs:
256
Run Growth:
-179
Growth Rate:
-69.92%
Total runs:
235
Run Growth:
-6.3K
Growth Rate:
-2697.02%
Total runs:
224
Run Growth:
-826
Growth Rate:
-368.75%
Total runs:
221
Run Growth:
221
Growth Rate:
100.00%
Total runs:
211
Run Growth:
-617
Growth Rate:
-292.42%
Total runs:
206
Run Growth:
-1.6K
Growth Rate:
-757.77%
Total runs:
194
Run Growth:
-85
Growth Rate:
-43.81%
Total runs:
193
Run Growth:
193
Growth Rate:
100.00%
Total runs:
172
Run Growth:
-207
Growth Rate:
-120.35%
Total runs:
161
Run Growth:
-938
Growth Rate:
-582.61%
Total runs:
140
Run Growth:
-402
Growth Rate:
-287.14%
Total runs:
137
Run Growth:
-768
Growth Rate:
-560.58%
Total runs:
132
Run Growth:
-320
Growth Rate:
-242.42%
Total runs:
131
Run Growth:
-602
Growth Rate:
-459.54%
Total runs:
128
Run Growth:
-300
Growth Rate:
-234.38%
Total runs:
128
Run Growth:
-859
Growth Rate:
-671.09%
Total runs:
124
Run Growth:
-224
Growth Rate:
-180.65%
Total runs:
117
Run Growth:
117
Growth Rate:
100.00%
Total runs:
117
Run Growth:
-877
Growth Rate:
-749.57%
Total runs:
117
Run Growth:
104
Growth Rate:
88.89%
Total runs:
116
Run Growth:
-733
Growth Rate:
-631.90%
Total runs:
112
Run Growth:
-900
Growth Rate:
-803.57%
Total runs:
112
Run Growth:
-780
Growth Rate:
-696.43%
Total runs:
111
Run Growth:
-1.0K
Growth Rate:
-909.91%
Total runs:
108
Run Growth:
-398
Growth Rate:
-368.52%
Total runs:
106
Run Growth:
-236
Growth Rate:
-222.64%
Total runs:
105
Run Growth:
-1.2K
Growth Rate:
-1109.52%
Total runs:
94
Run Growth:
82
Growth Rate:
87.23%
Total runs:
92
Run Growth:
-425
Growth Rate:
-461.96%
Total runs:
75
Run Growth:
16
Growth Rate:
21.33%
Total runs:
58
Run Growth:
-31
Growth Rate:
-53.45%
Total runs:
51
Run Growth:
-36
Growth Rate:
-70.59%
Total runs:
45
Run Growth:
-12
Growth Rate:
-26.67%
Total runs:
45
Run Growth:
-2
Growth Rate:
-4.44%