⚡ Each donation = another big MoE quantized
I host 30+ free APEX MoE quantizations as independent research. My only local hardware is an NVIDIA DGX Spark (122 GB unified memory), enough for ~30-50B-class MoEs, but bigger ones (200B+) require rented compute on H100/H200/Blackwell, typically $20-100 per quant.
If APEX quants are useful to you, your support directly funds those bigger runs.
🎉 Patreon (Monthly) | ☕ Buy Me a Coffee | ⭐ GitHub Sponsors
KAT-Coder-V2.5-Dev — APEX GGUF
APEX (Adaptive Precision for EXpert Models) quantizations of Kwaipilot/KAT-Coder-V2.5-Dev — Kwaipilot's Mixture-of-Experts model for agentic coding.
Brought to you by the LocalAI team | APEX Project | Technical Report
Available Files
| File | Profile | Best For |
|---|---|---|
| KAT-Coder-V2.5-Dev-APEX-I-Balanced.gguf | I-Balanced | Best overall — imatrix-enhanced |
| KAT-Coder-V2.5-Dev-APEX-I-Quality.gguf | I-Quality | Highest quality with imatrix |
| KAT-Coder-V2.5-Dev-APEX-Quality.gguf | Quality | Highest quality (no imatrix) |
| KAT-Coder-V2.5-Dev-APEX-Balanced.gguf | Balanced | General purpose |
| KAT-Coder-V2.5-Dev-APEX-I-Compact.gguf | I-Compact | Consumer GPUs, imatrix-enhanced |
| KAT-Coder-V2.5-Dev-APEX-Compact.gguf | Compact | Consumer GPUs |
| KAT-Coder-V2.5-Dev-APEX-I-Mini.gguf | I-Mini | Smallest viable, fastest inference |
What is APEX?
APEX is a quantization strategy for Mixture-of-Experts (MoE) models. It classifies tensors by role (routed expert, shared expert, attention) and applies a layer-wise precision gradient — edge layers (first/last 5) get higher precision, middle layers compress more aggressively. I-variants use diverse imatrix calibration (chat, code, reasoning, tool-calling, agentic traces, Wikipedia).
In MoE models the routed-expert FFN tensors dominate the weight budget but only ~8/256 experts activate per token, so APEX compresses middle-layer experts hardest while preserving edge layers, attention, and the always-active shared expert.
See the APEX project for full details.
Architecture
- Model: KAT-Coder-V2.5-Dev (Qwen3_5MoeForConditionalGeneration)
- Layers: 40 · Experts: 256 routed + 1 shared (8 active per token)
- Attention: 16 heads / 2 KV, hybrid (full attention every 4th layer)
- Calibration: v1.3 diverse dataset
Note: the config advertises an image token, but the released checkpoint ships no vision encoder weights, so these are text-only GGUFs (no mmproj).
Run with LocalAI
local-ai run mudler/KAT-Coder-V2.5-Dev-APEX-GGUF@KAT-Coder-V2.5-Dev-APEX-I-Balanced.gguf
Credits
APEX is brought to you by the LocalAI team. Built on llama.cpp. Base model by Kwaipilot.