Model Intelligence Sheet
mudler/gemma-4-26b-a4b-it-claude-opus-distill-apex-gguf overview
APEX (Adaptive Precision for EXpert Models) quantizations of gemma-4-26B-A4B-it-Claude-Opus-Distill — a Claude Opus reasoning-distilled version of google/gemma-4-26B-A4B-it by TeichAI. Brought to you by the LocalAI team | APEX Project | Technical Report
Downloads
14,410
Likes
11
Pipeline
—
Library
—
Visibility
Public
Access
Open
Repository Files & Downloads
8 files detected
Direct downloads for all repository files
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| gemma-4-26B-A4B-Claude-Distill-APEX-Balanced.gguf | GGUF | — | 18.17 GB | Download |
| gemma-4-26B-A4B-Claude-Distill-APEX-Compact.gguf | GGUF | — | 14.43 GB | Download |
| gemma-4-26B-A4B-Claude-Distill-APEX-I-Balanced.gguf | GGUF | — | 18.17 GB | Download |
| gemma-4-26B-A4B-Claude-Distill-APEX-I-Compact.gguf | GGUF | — | 14.43 GB | Download |
| gemma-4-26B-A4B-Claude-Distill-APEX-I-Mini.gguf | GGUF | — | 12.09 GB | Download |
| gemma-4-26B-A4B-Claude-Distill-APEX-I-Quality.gguf | GGUF | — | 19.16 GB | Download |
| gemma-4-26B-A4B-Claude-Distill-APEX-Quality.gguf | GGUF | — | 19.16 GB | Download |
| mmproj.gguf | GGUF | — | 1.11 GB | Download |
Model Details Live
Metadata Inspector
Normalized metadata (stored in metadata_json)
{
"metadata": {},
"card_data": {
"license": "gemma",
"base_model": "TeichAI/gemma-4-26B-A4B-it-Claude-Opus-Distill",
"tags": [
"gguf",
"quantized",
"apex",
"moe",
"mixture-of-experts",
"gemma4",
"claude-distilled",
"vlm",
"vision"
],
"frontmatter": {
"license": "gemma",
"base_model": "TeichAI/gemma-4-26B-A4B-it-Claude-Opus-Distill",
"tags": [
"gguf",
"quantized",
"apex",
"moe",
"mixture-of-experts",
"gemma4",
"claude-distilled",
"vlm",
"vision"
]
},
"hero_image_url": "",
"summary": "**APEX (Adaptive Precision for EXpert Models)** quantizations of gemma-4-26B-A4B-it-Claude-Opus-Distill — a Claude Opus reasoning-distilled version of google/gemma-4-26B-A4B-it by TeichAI. **Brought to you by the LocalAI team** | APEX Project | Technical Report",
"quick_links": [],
"benchmark_table_html": "",
"readme_markdown": "---\nlicense: gemma\nbase_model: TeichAI/gemma-4-26B-A4B-it-Claude-Opus-Distill\ntags:\n - gguf\n - quantized\n - apex\n - moe\n - mixture-of-experts\n - gemma4\n - claude-distilled\n - vlm\n - vision\n---\n\n# Gemma 4 26B-A4B Claude Opus Distill APEX GGUF\n\n**APEX (Adaptive Precision for EXpert Models)** quantizations of [gemma-4-26B-A4B-it-Claude-Opus-Distill](https://huggingface.co/TeichAI/gemma-4-26B-A4B-it-Claude-Opus-Distill) — a Claude Opus reasoning-distilled version of [google/gemma-4-26B-A4B-it](https://huggingface.co/google/gemma-4-26B-A4B-it) by TeichAI.\n\n**Brought to you by the [LocalAI](https://github.com/mudler/LocalAI) team** | [APEX Project](https://github.com/mudler/apex-quant) | [Technical Report](https://github.com/mudler/apex-quant/blob/main/paper/APEX_Technical_Report.pdf)\n\n## Benchmark Results\n\nBenchmarks coming soon. For reference APEX benchmarks on the Qwen3.5-35B-A3B architecture, see [mudler/Qwen3.5-35B-A3B-APEX-GGUF](https://huggingface.co/mudler/Qwen3.5-35B-A3B-APEX-GGUF).\n\n## What is APEX?\n\nAPEX 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 get higher precision, middle layers get more aggressive compression. I-variants use diverse imatrix calibration (chat, code, reasoning, tool-calling, agentic traces, Wikipedia).\n\nSee the [APEX project](https://github.com/mudler/apex-quant) for full details, technical report, and scripts.\n\n## Architecture\n\n- **Model**: gemma-4-26B-A4B-it-Claude-Opus-Distill (same architecture as gemma-4-26B-A4B-it)\n- **Layers**: 30\n- **Experts**: 128 routed (8 active per token)\n- **Total Parameters**: 26B\n- **Active Parameters**: ~4B per token\n- **Vision**: Built-in vision encoder (mmproj included)\n- **APEX Config**: 5+5 symmetric edge gradient across 30 layers\n- **Calibration**: v1.3 diverse dataset\n\n## Run with LocalAI\n\n```bash\nlocal-ai run mudler/gemma-4-26B-A4B-it-Claude-Opus-Distill-APEX-GGUF@gemma-4-26B-A4B-Claude-Distill-APEX-I-Balanced.gguf\n```\n\n## Credits\n\nAPEX is brought to you by the [LocalAI](https://github.com/mudler/LocalAI) team. Developed through human-driven, AI-assisted research. Built on [llama.cpp](https://github.com/ggerganov/llama.cpp).\n",
"related_quantizations": []
},
"tags": [
"gguf",
"quantized",
"apex",
"moe",
"mixture-of-experts",
"gemma4",
"claude-distilled",
"vlm",
"vision",
"base_model:TeichAI/gemma-4-26B-A4B-it-Claude-Opus-Distill",
"base_model:quantized:TeichAI/gemma-4-26B-A4B-it-Claude-Opus-Distill",
"license:gemma",
"endpoints_compatible",
"region:us",
"conversational"
],
"likes": 11,
"downloads": 14410,
"gated": false,
"private": false,
"last_modified": "2026-04-05T11:01:01.000Z",
"created_at": "2026-04-05T02:57:32.000Z",
"pipeline_tag": "",
"library_name": ""
}
Source payload excerpt (from Hugging Face API)
{
"_id": "69d1cf9c8b740d944df1e996",
"id": "mudler/gemma-4-26B-A4B-it-Claude-Opus-Distill-APEX-GGUF",
"modelId": "mudler/gemma-4-26B-A4B-it-Claude-Opus-Distill-APEX-GGUF",
"sha": "73a81f667f2aea0e3b56149a2e829acc28666001",
"createdAt": "2026-04-05T02:57:32.000Z",
"lastModified": "2026-04-05T11:01:01.000Z",
"author": "mudler",
"downloads": 14410,
"likes": 11,
"gated": false,
"private": false,
"pipeline_tag": "",
"library_name": "",
"siblings_count": 10
}