Linkbricks Horizon AI Korean Advanced 56B is an open-source language model by Saxo. Features: 46.7b LLM, VRAM: 93.6GB, Context: 32K, License: apache-2.0, MoE, Quantized, LLM Explorer Score: 0.27, ELO: 1084, Arc: 71.4, HellaSwag: 87.2, MMLU: 72.3, GSM8K: 70.7.
Base model:nousresearch/nous-h... Base model:quantized:nousresea... Cn Conversational Dataset:kuotient/gsm8k-ko Dataset:lilacai/glaive-functio... Dataset:maywell/ko ultrafeedba... Dataset:saxo/en ko translation... Dataset:saxo/en ko translation... Dataset:saxo/ko-news-corpus-1 Dataset:saxo/ko-news-corpus-2 Dataset:saxo/ko-news-corpus-3 Dataset:saxo/ko-news-corpus-4 Dataset:saxo/ko-news-corpus-5 Dataset:saxo/ko-news-corpus-6 Dataset:saxo/ko-news-corpus-7 Dataset:saxo/ko-news-corpus-8 Dataset:saxo/ko-news-corpus-9 Dataset:saxo/ko aspect sentime... Dataset:saxo/ko cn translation... Dataset:saxo/ko government qa ... Dataset:saxo/ko jp translation... Dataset:saxo/ko summarization ... Dataset:saxo/openorca cleaned ... Dataset:youjunhyeok/ko-orca-pa... En Endpoints compatible Gguf Jp Ko Mixtral Moe Quantized Region:us Safetensors Sharded Tensorflow
Model Card on HF 🤗: https://huggingface.co/Saxo/Linkbricks-Horizon-AI-Korean-Advanced-56B
Linkbricks Horizon AI Korean Advanced 56B Benchmarks
LMSys ELO: 1084 vs 1272 (so35)-14.8%
ARC: 71.42 vs 96.7 (so35)-26.1%
HellaSwag: 87.21 vs 95.3 (gpt4)-8.5%
MMLU: 72.28 vs 88.3 (so35)-18.1%
GSM8K: 70.66 vs 96.4 (so35)-26.7%
LLME Score: 0.26759
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
What is the LLM Explorer Rank (Score)
Linkbricks Horizon AI Korean Advanced 56B Parameters and Internals
| LLM Name | Linkbricks Horizon AI Korean Advanced 56B |
| Repository 🤗 | https://huggingface.co/Saxo/Linkbricks-Horizon-AI-Korean-Advanced-56B |
| Base Model(s) | |
| Model Size | 46.7b |
| Required VRAM | 93.6 GB |
| Updated | 2026-05-11 |
| Maintainer | Saxo |
| Model Type | mixtral |
| Model Files | |
| Supported Languages | ko en jp cn |
| GGUF Quantization | Yes |
| Quantization Type | gguf |
| Model Architecture | MixtralForCausalLM |
| License | apache-2.0 |
| Context Length | 32768 |
| Model Max Length | 32768 |
| Transformers Version | 4.43.2 |
| Tokenizer Class | LlamaTokenizer |
| Padding Token | <|im_end|> |
| Vocabulary Size | 32002 |
| Torch Data Type | bfloat16 |
Best Alternatives to Linkbricks Horizon AI Korean Advanced 56B
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Mixtral 8x7B Instruct V0.1 | 32K / 93.6 GB | 774907 | 4683 |
| Nous Hermes 2 Mixtral 8x7B DPO | 32K / 93.6 GB | 9047 | 453 |
| Mixtral 8x7B V0.1 | 32K / 93.6 GB | 112078 | 1808 |
| Sensualize Mixtral Bf16 | 32K / 93.6 GB | 0 | 0 |
| Skadi Mixtral V1 | 32K / 93.5 GB | 0 | 0 |
| Mixtral 8x7B Instruct V0.1 FP8 | 32K / 47.1 GB | 112011 | 0 |
| Franziska Mixtral V1 | 32K / 93.5 GB | 0 | 0 |
| Typhon Mixtral V1 | 32K / 93.4 GB | 0 | 0 |
| GritLM 8x7B KTO | 32K / 93.6 GB | 7952 | 3 |
| Smaug Mixtral V0.1 | 32K / 187.7 GB | 8548 | 12 |
Note: green Score (e.g. "73.2") means that the model is better than Saxo/Linkbricks-Horizon-AI-Korean-Advanced-56B .
Rank the Linkbricks Horizon AI Korean Advanced 56B Capabilities
🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟
What open-source LLMs or SLMs are you in search of? 53972 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a