EL

ELYZA Thinking 1.0 llm jp 4 32b a3b

ElyzaGGUF by na0x2c6Updated 2026-10-03
Download on Hugging Face
Parameters32B
Context64K
Layers32
Architectureqwen3moe
Smallest quant0.1 GB
Quick answer

ELYZA Thinking 1.0 llm jp 4 32b a3b needs about 22.4 GB of memory at Q4_K_M (a 21.4 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 10 quants, from 0.1 GB to 64.3 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 2.1 GB.

32B Japanese reasoning model with MoE

ELYZA Thinking 1.0 llm jp 4 32b a3b is a 32‑billion‑parameter MoE model trained for Japanese and English reasoning. It supports 65536‑token context and is licensed under Apache‑2.0.

Good at

  • Japanese reasoning
  • Instruction following
  • Math problem solving

Limits

  • Requires patched llama.cpp
  • Sensitive to long blank lines
  • No vision support

Best for

  • Japanese QA systems
  • Educational tutoring
  • Research on Japanese language models

Who it’s for. Developers building Japanese‑centric LLM applications or researchers needing a large MoE model for reasoning tasks. The model is open‑source under Apache‑2.0 and can be run locally with llama.cpp.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.1 GB 1.1 GB
Fits Get file →
Q3_K_M16.7 GB 17.8 GB
Too big Get file →
IQ4_XS17.6 GB 18.6 GB
Too big Get file →
Q3_K17.8 GB 18.8 GB
Too big Get file →
IQ4_XS20.5 GB 21.5 GB
Too big Get file →
Q4_K_M21.4 GB 22.4 GB
Too big Get file →
Q5_K_M24.4 GB 25.4 GB
Too big Get file →
Q6_K28.8 GB 29.9 GB
Too big Get file →
Q8_034.2 GB 35.2 GB
Too big Get file →
BF1664.3 GB (2 parts) 65.3 GB
Too big Get file →

For 8 GB VRAM, use Q4_K_M (21.4 GB file, 22.4 GB memory at 8k context). For 12 GB VRAM, Q5_K_M (24.41 GB file, 25.4 GB memory). For 16 GB VRAM, Q6_K (28.84 GB file, 29.9 GB memory).

What GPU do you need for ELYZA Thinking 1.0 llm jp 4 32b a3b?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: the best quant that fits is the main file (0.1 GB file, about 1.1 GB in use).
  • 6 GB VRAM: the best quant that fits is the main file (0.1 GB file, about 1.1 GB in use).
  • 8 GB VRAM: the best quant that fits is the main file (0.1 GB file, about 1.1 GB in use).
  • 12 GB VRAM: the best quant that fits is the main file (0.1 GB file, about 1.1 GB in use).
  • 16 GB VRAM: the best quant that fits is the main file (0.1 GB file, about 1.1 GB in use).
  • 24 GB VRAM: the best quant that fits is Q4_K_M (21.4 GB file, about 22.4 GB in use).
  • 32 GB VRAM: the best quant that fits is Q6_K (28.8 GB file, about 29.9 GB in use).
  • 48 GB VRAM: the best quant that fits is Q8_0 (34.2 GB file, about 35.2 GB in use).

Which ELYZA Thinking 1.0 llm jp 4 32b a3b quant should you choose?

Q4_K_M is the usual starting point for ELYZA Thinking 1.0 llm jp 4 32b a3b: it balances file size and quality well. Here it is a 21.4 GB file.

The Q8_0 file is 34.2 GB, about 1.6 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

If memory is tight, the smallest file is the first one in the table at 0.1 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen3moe
Parameters32B (approximate, mixture of experts)
Layers32
Trained context65,536 tokens
KV cache (fp16)64 KB per token

Source

Quantized byna0x2c6
FamilyElyza
Licenseapache-2.0
Downloads (HF)456
Last updated2026-10-03

Frequently asked questions about ELYZA Thinking 1.0 llm jp 4 32b a3b

How much VRAM does ELYZA Thinking 1.0 llm jp 4 32b a3b need?

About 22.4 GB at Q4_K_M with an 8K context window. That is the 21.4 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run ELYZA Thinking 1.0 llm jp 4 32b a3b on an 8 GB GPU?

Yes, with the smaller files: the 0.1 GB file needs about 1.1 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of ELYZA Thinking 1.0 llm jp 4 32b a3b GGUF?

It depends on the quant: file is 0.1 GB, Q3_K_M is 16.7 GB, IQ4_XS is 17.6 GB, Q3_K is 17.8 GB, IQ4_XS is 20.5 GB, Q4_K_M is 21.4 GB and 4 more.

What context length does ELYZA Thinking 1.0 llm jp 4 32b a3b support?

The GGUF header lists a trained context length of 65,536 tokens (64K). The KV cache grows with the context you use: about 0.5 GB at 8K and 2.1 GB at 32K.

Where can I download ELYZA Thinking 1.0 llm jp 4 32b a3b GGUF?

The files are on Hugging Face in the na0x2c6/ELYZA-Thinking-1.0-llm-jp-4-32b-a3b-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does ELYZA Thinking 1.0 llm jp 4 32b a3b use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run this model on 8 GB VRAM?

Use the Q4_K_M quantized file; it requires about 22.4 GB memory at 8k context, so 8 GB VRAM is insufficient.

Does this model support vision inputs?

No, the model is text‑only and has no vision capabilities.

What patch is needed for llama.cpp?

Apply scripts/llmjp4-space-after-control.patch and rebuild llama.cpp to enable chat completions.

Other GGUFs of ELYZA-Thinking-1.0-llm-jp-4-32b-a3b

Similar models

Browse all →

More from na0x2c6

All by na0x2c6 →

More Elyza models

All Elyza →

Other medium models

All medium models →