[whatmodel.sh](../../README.md) / KAT-Coder-V2.5-Dev (35B-A3B)
# KAT-Coder-V2.5-Dev (35B-A3B)
Kuaishou (Kwaipilot) · released 2026-07 · 35B params (3B active) · apache-2.0
**score 69.4** ±1.5 · #85 of 732 · from 5 benchmark results across 1 source · † no independently run results verified
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks KAT-Coder-V2.5-Dev (35B-A3B) among every model on that task or test, worst to best.
task vs the others rank counts tests
agent tasks ▁▁▃▆█▄▅●▄▄▂▁ #88 of 313 87% 4
coding ▁▁▁▃▅▇█▇▅●▃▁ #71 of 469 72% 4
agent tasks · #88 of 313 · 87% of its score
Beats [Grok 4.20 (0309 reasoning)](grok-4-20.md), [GPT-5.1](gpt-5-1.md), [o3-pro](o3-pro.md) here, though they rank higher overall. Trails [MiniMax-M2.7](minimax-m2-7.md), [Step 3.5 Flash (196B-A11B)](step-3-5-flash.md), [Nemotron 3.5 Bolt 74B A9B](nemotron-3-5-bolt-74b-a9b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
Terminal-Bench 2.1 (maker) 32.6% ▄······█▄●█▄ beats 4/7 very hard · avg 7.2% 27% sharp · at its level
SWE-bench Verified (maker) 69.4% ▄▄·····▄▄▄▄● beats 7/7 hard · avg 16% 24% sharp · at its level
SWE-bench Pro (split unspecified)… 46.0% ▄···▄···▄▄█● beats 7/7 very hard · avg 6.2% 23% sharp · at its level
SWE-bench Multilingual (maker) 63.0% ▄·▄·····█▄█● beats 7/7 hard · avg 18% 13% at its level
coding · #71 of 469 · 72% of its score
Beats [DeepSeek V4 Flash Vision (Exp)](deepseek-v4-flash-vision-exp.md), [GPT-5.2](gpt-5-2.md), [GPT-5](gpt-5.md) here, though they rank higher overall. Trails [MiMo-V2.5-Pro (1T-A42B)](mimo-v2-5-pro.md), [Qwen3.6 Plus](qwen3-6-plus.md), [MiniMax-M3](minimax-m3.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
SWE-bench Verified (maker) 69.4% ▄▄·····▄▄▄▄● beats 7/7 hard · avg 16% 24% sharp · at its level
SWE-bench Pro (split unspecified)… 46.0% ▄···▄···▄▄█● beats 7/7 very hard · avg 6.2% 23% sharp · at its level
SWE-bench Multilingual (maker) 63.0% ▄·▄·····█▄█● beats 7/7 hard · avg 18% 13% at its level
SciCode (level unspecified) (make… 44.2% ▄···▄▄█▄·▄·● beats 7/7 very hard · avg 13% 12% at its level
source results share of the score
Kwaipilot/KAT-Coder-V2.5-Dev model card… 5 99% ▇▇▇▇▇▇▇▇▇▇▇▇
prior: pull toward the middle 1% ▇
weighted average = 69.4
Licences: Kwaipilot/KAT-Coder-V2.5-Dev model card ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0))
## How to run it
### llama.cpp (recommended)
1. Install llama.cpp. Homebrew and winget installs name the commands llama-cli and llama-server instead. [Other ways to install](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md)
```sh
# macOS / Linux
curl -LsSf https://llama.app/install.sh | sh
# Windows
irm https://llama.app/install.ps1 | iex
```
2. Run KAT-Coder-V2.5-Dev (35B-A3B) (Q4_K_M, needs ~21.39 GB, from [huggingface.co/bartowski](https://huggingface.co/bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF/tree/d8f684f08d2950ea9d2db6a35ef7dada0707858b))
```sh
llama cli -hf bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF:Q4_K_M
```
Other versions: Q8_0 (~36.91 GB) `llama cli -hf bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF:Q8_0`
### LM Studio
1. Install LM Studio. Open the app once before using the lms command. [Download it (macOS / Windows / Linux)](https://lmstudio.ai/download)
2. Run KAT-Coder-V2.5-Dev (35B-A3B) (Q4_K_M, needs ~21.39 GB, from [huggingface.co/bartowski](https://huggingface.co/bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF/tree/d8f684f08d2950ea9d2db6a35ef7dada0707858b))
```sh
lms get https://huggingface.co/bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF@Q4_K_M -y
```
3. Then run lms ls to see its name, and lms chat followed by that name
Other versions: Q8_0 (~36.91 GB) `lms get https://huggingface.co/bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF@Q8_0 -y` · 4bit (~19.5 GB, Mac only) `lms get https://huggingface.co/majentik/KAT-Coder-V2.5-Dev-MLX-4bit -y` · 8bit (~36.8 GB, Mac only) `lms get https://huggingface.co/majentik/KAT-Coder-V2.5-Dev-MLX-8bit -y`
### Ollama
1. Install Ollama. [Download it (macOS / Windows)](https://ollama.com/download)
```sh
# Linux
curl -fsSL https://ollama.com/install.sh | sh
```
2. Run KAT-Coder-V2.5-Dev (35B-A3B) (Q4_K_M, needs ~21.39 GB, from [huggingface.co/bartowski](https://huggingface.co/bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF/tree/d8f684f08d2950ea9d2db6a35ef7dada0707858b))
```sh
ollama run hf.co/bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF:Q4_K_M
```
Other versions: Q8_0 (~36.91 GB) `ollama run hf.co/bartowski/Kwaipilot_KAT-Coder-V2.5-Dev-GGUF:Q8_0`
### MLX (Mac)
1. Install MLX (Mac). Apple silicon Macs only, text only.
```sh
# with uv
uv tool install mlx-lm
# or
pip install mlx-lm
```
2. Run KAT-Coder-V2.5-Dev (35B-A3B) (4bit, needs ~19.5 GB, Mac only, from [huggingface.co/majentik](https://huggingface.co/majentik/KAT-Coder-V2.5-Dev-MLX-4bit/tree/526fa15b962e0e8bf40ed744f4fdc39e6679b1e6))
```sh
mlx_lm.chat --model majentik/KAT-Coder-V2.5-Dev-MLX-4bit
```
3. Or start an OpenAI-compatible server at http://localhost:8080/v1
```sh
mlx_lm.server --model majentik/KAT-Coder-V2.5-Dev-MLX-4bit
```
Other versions: 8bit (~36.8 GB, Mac only) `mlx_lm.chat --model majentik/KAT-Coder-V2.5-Dev-MLX-8bit`
Model repository and documentation: [huggingface.co/Kwaipilot](https://huggingface.co/Kwaipilot/KAT-Coder-V2.5-Dev)
[interactive version](https://whatmodel.sh/model/kat-coder-v2-5-dev) · [JSON](https://whatmodel.sh/api?format=json&all=1)