Hugging Face's logo

TokenRhythm
/
NeoHorse-1-4B-MLX

Text Generation
MLX
Safetensors
qwen3_5
agentic
tool-use
coding
reasoning
instruction-following
bf16
conversational
Model card Files
xet
 Community
NeoHorse-1-4B-MLX
File size: 26,476 Bytes
33bde7c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
---
license: apache-2.0
library_name: mlx
pipeline_tag: text-generation
base_model: TokenRhythm/NeoHorse-1-4B
tags:
- agentic
- tool-use
- coding
- reasoning
- instruction-following
- mlx
- bf16
---

## MLX local inference

This is the **unquantized BF16 MLX** version of [NeoHorse-1-4B](https://huggingface.co/TokenRhythm/NeoHorse-1-4B) for Apple Silicon. Converted from the original BF16 weights with MLX-LM. No weight quantization is applied; MLX-LM adapts tensor names/layouts and normalization representation for its runtime. Benchmark scores below refer to the original model, not a separate evaluation of this MLX version.

```bash
pip install "mlx-lm>=0.31.3"
mlx_lm.chat --model TokenRhythm/NeoHorse-1-4B-MLX
```

The model downloads automatically from Hugging Face. The original chat template is preserved. See [Deployment](#deployment) for local checkpoints, the chat API, and tool calling.

<div align="center">
  <h1>NeoHorse-1-4B</h1>
  <p><b>Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness.</b></p>
</div>

<div align="center">
  <a href="https://github.com/TokenRhythm/NeoHorse"><img alt="GitHub" src="https://img.shields.io/badge/GitHub-NeoHorse-181717?logo=github&logoColor=white"></a>
  <a href="https://tokenrhythm.ai/"><img alt="Company" src="https://img.shields.io/badge/Company-TokenRhythm-F97316?logo=homeassistant&logoColor=white"></a>
  <a href="https://huggingface.co/TokenRhythm"><img alt="Hugging Face" src="https://img.shields.io/badge/Hugging%20Face-Models-FFD21E?logo=huggingface&logoColor=000000"></a>
  <a href="https://x.com/opensquilla"><img alt="Twitter / X" src="https://img.shields.io/badge/Twitter%20%2F%20X-OpenSquilla-111827?logo=x&logoColor=white"></a>
  <a href="https://www.apache.org/licenses/LICENSE-2.0"><img alt="License: Apache-2.0" src="https://img.shields.io/badge/License-Apache--2.0-64748B"></a>
</div>

<p align="center">
  <a href="https://arxiv.org/abs/2609.08183"><b>Technical Report</b></a>
</p>

<style>
/* Reusable benchmark table architecture. Inline styles remain as a fallback for HF rendering. */
.vl-table {
  width: 100%;
  min-width: 100%;
  border-collapse: collapse;
  table-layout: fixed;
  font-size: 15px;
}
.vl-table th {
  font-size: 15px !important;
  line-height: 1.2;
  color: #c2410c;
  background: rgba(249,115,22,.10);
}
.vl-table td:not(.benchmark-cell):not([colspan]) {
  font-size: 15px;
  line-height: 1.2;
  vertical-align: middle;
}
.vl-table .benchmark-cell {
  padding: 12px 10px 12px 18px !important;
  vertical-align: middle;
}
.vl-table .benchmark-capability {
  font-size: 15px;
  font-weight: 600;
  line-height: 1.22;
  color: #c2410c;
}
.vl-table .benchmark-name {
  margin-top: 4px;
  font-size: 11px;
  font-weight: 400;
  line-height: 1.2;
  color: inherit;
}
.vl-table .metric-stack {
  display: flex;
  flex-direction: column;
  gap: 7px;
  padding: 3px 0;
}
.vl-table .metric-label {
  font-size: 10px;
  font-weight: 400;
  line-height: 1.1;
  color: inherit;
}
.vl-table .metric-value {
  margin-top: 2px;
  font-size: 15px;
  line-height: 1.15;
  color: inherit;
}
.model-table td:first-child {
  width: 34%;
  font-weight: 600;
}
/* HF's theme toggle sets the dark class on an ancestor. */
.dark .vl-table th,
.dark .vl-table .benchmark-capability {
  color: #fdba74 !important;
}
</style>

NeoHorse-1-4B is a 4B causal language model and an initial prototype on the path toward **recursive self-improvement (RSI)**. It is post-trained from Qwen3.5-4B for text-based agent harnesses, tool use, coding, and instruction following.

Derived from [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B) and fine-tuned by TokenRhythm. The source checkpoint was repackaged for text-only inference. This repository contains **language-model weights only**, converted to MLX BF16 without weight quantization.

<p align="center">
  <a href="https://huggingface.co/TokenRhythm/NeoHorse-1-4B/resolve/main/4B_head_fig.jpg">
    <img src="https://huggingface.co/TokenRhythm/NeoHorse-1-4B/resolve/main/4B_head_fig.jpg" alt="NeoHorse-1-4B evaluation results" width="100%">
  </a>
</p>

## Highlights

- **Path toward RSI:** the routing harness assigns tasks to a heterogeneous model pool, records tool interactions and outcomes, estimates capability demand, and uses capability-level feedback to shape the next training mixture. Updated models can return to the harness, closing a prototype evaluation鈥搒election鈥搖pdate loop; extending this loop across successive iterations is the next step toward RSI.
- **Agentic post-training framework:** the associated research explores routing-guided curriculum SFT and routing-guided on-policy distillation to turn execution trajectories into training signal while preserving execution and harness context around each response.
- **Data quality:** exact and near-duplicate removal, evaluation decontamination, structural validation, six-dimensional semantic evaluation, and subscene-level Scene/Goal/Outcome labeling.
- **Broad gains:** 64.87 macro average across ten benchmarks versus 58.94 for Qwen3.5-4B (**+5.93**).

## Model Details

<div style="width:100%;max-width:none;margin:16px 0;padding:0;overflow-x:auto">
<table class="vl-table model-table" width="100%" style="display:table;width:100%;min-width:100%;table-layout:fixed;border-collapse:collapse;font-size:13px">
<thead><tr>
<th style="padding:9px 10px;text-align:left;border-bottom:2px solid #f97316;color:#c2410c;background:rgba(249,115,22,.10)">Property</th>
<th style="padding:9px 10px;text-align:left;border-bottom:2px solid #f97316;color:#c2410c;background:rgba(249,115,22,.10)">Value</th>
</tr></thead><tbody>
<tr>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16);font-weight:600">Model family</td>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16)">NeoHorse Agent-Native Causal Language Model</td>
</tr>
<tr>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16);font-weight:600">Parameters</td>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16)">Approximately <strong>4B</strong></td>
</tr>
<tr>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16);font-weight:600">Base model</td>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16)"><a href="https://huggingface.co/Qwen/Qwen3.5-4B">Qwen3.5-4B</a></td>
</tr>
<tr>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16);font-weight:600">Post-training</td>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16)">Routing-guided agentic post-training</td>
</tr>
<tr>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16);font-weight:600">Interface</td>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16)">Text input and text output</td>
</tr>
<tr>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16);font-weight:600">Context length</td>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16)">262,144 natively and extensible up to 1,010,000 tokens.</td>
</tr>
<tr>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16);font-weight:600">Weight format / precision</td>
<td style="padding:9px 10px;border-bottom:1px solid rgba(249,115,22,.16)">MLX Safetensors / BF16 (unquantized)</td>
</tr>
</tbody></table>
</div>

## Evaluation
The 4B track compares NeoHorse-1-4B with five representative open-weight models. Results are grouped by capability in the table below. Higher is better; `螖` is NeoHorse-1-4B minus Qwen3.5-4B. **Bold** marks the best available result; <ins>underlining</ins> marks the second-best.

<div style="overflow-x:auto">
<table class="vl-table" width="100%" style="display:table;width:100%;min-width:100%;border-collapse:collapse;table-layout:fixed;font-size:13px">
<thead><tr>
<th style="padding:9px 8px;text-align:center;border-bottom:2px solid #f97316;color:#c2410c;background:rgba(249,115,22,.10);text-align:left">Benchmark</th>
<th style="padding:9px 8px;text-align:center;border-bottom:2px solid #f97316;color:#c2410c;background:rgba(249,115,22,.10)">Qwen3.5-4B</th>
<th style="padding:9px 8px;text-align:center;border-bottom:2px solid #f97316;color:#c2410c;background:rgba(249,115,22,.10)">Gemma-4-E4B-it</th>
<th style="padding:9px 8px;text-align:center;border-bottom:2px solid #f97316;color:#c2410c;background:rgba(249,115,22,.10)">Nanbeige-4.2-3B</th>
<th style="padding:9px 8px;text-align:center;border-bottom:2px solid #f97316;color:#c2410c;background:rgba(249,115,22,.10)">Agents-A1-4B</th>
<th style="padding:9px 8px;text-align:center;border-bottom:2px solid #f97316;color:#c2410c;background:rgba(249,115,22,.10)">Spark-X2.5-4B</th>
<th style="padding:9px 8px;text-align:center;border-bottom:2px solid #f97316;color:#c2410c;background:rgba(249,115,22,.10);background:rgba(249,115,22,.18)">NeoHorse-1-4B</th>
<th style="padding:9px 8px;text-align:center;border-bottom:2px solid #f97316;color:#c2410c;background:rgba(249,115,22,.10);background:rgba(249,115,22,.18)">螖 vs Qwen3.5-4B</th>
</tr></thead><tbody>
<tr><td class="benchmark-capability" colspan="8" style="padding:10px 8px;font-weight:700;color:#c2410c;background:rgba(249,115,22,.10);border-top:2px solid #f97316">馃 Agentic</td></tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">QwenClawBench</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">38.47</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">22.98</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">40.66</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">43.16</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><ins>43.52</ins></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value"><strong>44.68</strong></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+6.21</span></div></td>
</tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">WorkBuddy Bench</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">24.62</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">11.65</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">21.03</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><ins>33.37</ins></span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">26.47</span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value"><strong>34.41</strong></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+9.79</span></div></td>
</tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">PinchBench</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">71.19</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">47.60</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">66.78</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><ins>75.07</ins></span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">62.37</span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value"><strong>77.33</strong></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+6.14</span></div></td>
</tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">VitaBench</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">21.50</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">5.00</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">31.50</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><strong>39.25</strong></span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><ins>37.00</ins></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">32.00</span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+10.50</span></div></td>
</tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">BFCL v4</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">61.02</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">47.18</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><strong>67.28</strong></span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">46.60</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><ins>63.71</ins></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">61.79</span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+0.77</span></div></td>
</tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">tau2-Bench</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">84.29</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">43.60</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><ins>85.08</ins></span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">81.00</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">77.72</span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value"><strong>88.46</strong></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+4.17</span></div></td>
</tr>
<tr><td class="benchmark-capability" colspan="8" style="padding:10px 8px;font-weight:700;color:#c2410c;background:rgba(249,115,22,.10);border-top:2px solid #f97316">馃捇 Coding</td></tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">HumanEval</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">87.20</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">84.76</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><strong>98.78</strong></span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">92.68</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">92.07</span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value"><ins>96.95</ins></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+9.75</span></div></td>
</tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">LiveCodeBench v6</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">53.71</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">52.00</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><strong>72.50<sup>*</sup></strong></span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">56.57</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">54.86</span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value"><ins>59.43</ins></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+5.72</span></div></td>
</tr>
<tr><td class="benchmark-capability" colspan="8" style="padding:10px 8px;font-weight:700;color:#c2410c;background:rgba(249,115,22,.10);border-top:2px solid #f97316">馃摎 Instruction Following</td></tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">IFBench</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">60.33</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">40.00</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">55.00</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">63.33</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><strong>73.33</strong></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value"><ins>65.33</ins></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+5.00</span></div></td>
</tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">IFEval</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">87.06</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">74.68</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">84.47</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">83.55</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><strong>91.13</strong></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value"><ins>88.35</ins></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+1.29</span></div></td>
</tr>
<tr><td class="benchmark-capability" colspan="8" style="padding:10px 8px;font-weight:700;color:#c2410c;background:rgba(249,115,22,.10);border-top:2px solid #f97316">馃搳 Overall</td></tr>
<tr style="border-bottom:1px solid rgba(128,128,128,.16)">
<td class="benchmark-cell" style="padding:8px;font-weight:600"><div class="benchmark-name">Ten-benchmark average</div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">58.94</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">42.95</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value"><ins>62.31</ins></span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">61.46</span></div></td>
<td style="padding:8px;text-align:center"><div class="metric-stack"><span class="metric-value">62.22</span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value"><strong>64.87</strong></span></div></td>
<td style="padding:8px;text-align:center;background:rgba(249,115,22,.16);font-weight:700"><div class="metric-stack"><span class="metric-value">+5.93</span></div></td>
</tr>
</tbody></table>
</div>

`*` Nanbeige-4.2-3B LiveCodeBench v6 result is reported in the corresponding model's official blog post or technical report.

> **Reported protocol:** SGLang v0.5.17 路 `temperature=1.0` 路 `top_p=0.95` 路 `top_k=20` 路 `min_p=0.0` 路 `presence_penalty=1.5` 路 `repetition_penalty=1.0` 路 thinking mode enabled with `enable_thinking=true` and `force_nonempty_content=true`. QwenClawBench, WorkBuddy Bench, and tau2-Bench use three runs; PinchBench and VitaBench use one run; the remaining benchmarks follow their official protocols. VitaBench uses the DeepSeek-V4-Flash simulator and judge.

## Deployment

Use [MLX-LM](https://github.com/ml-explore/mlx-lm) on an Apple Silicon Mac to run this checkpoint.

### Install and select a local checkpoint

```bash
pip install "mlx-lm>=0.31.3"
MODEL_PATH="/path/to/NeoHorse-1-4B-MLX"
```

Set `MODEL_PATH` to the downloaded MLX directory containing `config.json`, tokenizer files, `chat_template.jinja`, and model weights. You can also use `TokenRhythm/NeoHorse-1-4B-MLX` as the model path to download it automatically from Hugging Face.

### Chat locally

```bash
mlx_lm.chat --model "$MODEL_PATH"
```

### Start an API server

```bash
mlx_lm.server \
  --model "$MODEL_PATH" \
  --host 127.0.0.1 \
  --port 8080
```

The server exposes an OpenAI-compatible `/v1/chat/completions` endpoint. In the requests below, `default_model` refers to the checkpoint selected with `--model`.

### Basic Usage

After the server starts, run this request in another terminal:

```bash
curl http://127.0.0.1:8080/v1/chat/completions \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "default_model",
    "messages": [
      {"role": "user", "content": "Write a Python function that returns the first n Fibonacci numbers."}
    ],
    "max_tokens": 2048,
    "stream": false
  }'
```

The generated reply is returned in `choices[0].message.content`.

### Tool Calling

Pass function definitions in the `tools` field:

```bash
curl http://127.0.0.1:8080/v1/chat/completions \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "default_model",
    "messages": [
      {"role": "user", "content": "Use get_weather to check the current weather in Beijing in celsius."}
    ],
    "tools": [
      {
        "type": "function",
        "function": {
          "name": "get_weather",
          "description": "Get the current weather for a city.",
          "parameters": {
            "type": "object",
            "properties": {
              "city": {"type": "string", "description": "City name."},
              "unit": {"type": "string", "enum": ["celsius", "fahrenheit"]}
            },
            "required": ["city", "unit"]
          }
        }
      }
    ],
    "max_tokens": 2048,
    "stream": false
  }'
```

MLX-LM reads the preserved chat template to format tool requests and parse generated calls. When the model chooses to call a tool, the call is returned in `choices[0].message.tool_calls`. Your application executes the function, appends the assistant message and a `role: "tool"` result with the matching `tool_call_id`, then sends the conversation back to the same endpoint for the final answer.

## License

NeoHorse-1-4B is released under the **Apache License 2.0**.

The upstream model is [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B). Its original copyright notice, Copyright 2026 Alibaba Cloud, is retained in the license file. TokenRhythm fine-tuned and repackaged the source checkpoint for text-only inference. This repository provides its MLX BF16 conversion without weight quantization.

## Citation

```
@misc{neohorse2026,
  title        = {NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness},
  author       = {NeoHorse Team},
  year         = {2026},
  howpublished = {arXiv preprint},
  eprint       = {2609.08183},
  archivePrefix = {arXiv},
  primaryClass = {cs.CL},
  url          = {https://arxiv.org/abs/2609.08183}
}
```

For questions or issue reports, use the [NeoHorse project repository](https://github.com/TokenRhythm/NeoHorse).