KI

Kimi K2.5

KimiGGUF by bartowskiUpdated 2026-02-17
Download on Hugging Face
Parameters1030B
Context256K
Layers61
Architecturedeepseek2
Smallest quant1.5 GB
Quick answer

Kimi K2.5 needs about 581.9 GB of memory at Q4_K_M (a 580.4 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 27 quants, from 1.5 GB to 643.2 GB (Q4_1). Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.

Kimi K2.5 – multimodal agentic model with vision and language

Kimi K2.5 is a 1.03 B parameter model built on the deepseek2 architecture. It supports vision, language, and agentic tool use. The model is trained on 262 144 token context length and is available in many GGUF quantizations.

Good at

  • Vision‑language reasoning
  • Agentic tool orchestration
  • Code generation from visual specs

Limits

  • No text‑only mode, requires vision input
  • Large VRAM needed for high‑quality quants
  • License is other, not open source

Best for

  • Visual data analysis
  • Multimodal chatbot development
  • Automated UI code generation

Who it’s for. Developers building multimodal applications that need both vision and language understanding. Researchers exploring agentic behavior in LLMs may also find it useful.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—1.5 GB 3.1 GB
Fits Get file →
IQ1_S210.3 GB (6 parts) 211.9 GB
Too big Get file →
IQ1_M219.6 GB (6 parts) 221.2 GB
Too big Get file →
IQ2_XXS245.3 GB (7 parts) 246.9 GB
Too big Get file →
IQ2_XS283.1 GB (8 parts) 284.7 GB
Too big Get file →
IQ2_S283.8 GB (8 parts) 285.4 GB
Too big Get file →
IQ2_M323.0 GB (9 parts) 324.5 GB
Too big Get file →
Q2_K358.6 GB (10 parts) 360.2 GB
Too big Get file →
Q2_K_L359.8 GB (10 parts) 361.4 GB
Too big Get file →
IQ3_XXS404.6 GB (11 parts) 406.2 GB
Too big Get file →
IQ3_XS420.1 GB (11 parts) 421.7 GB
Too big Get file →
Q3_K_S444.8 GB (12 parts) 446.4 GB
Too big Get file →
IQ3_M466.9 GB (13 parts) 468.5 GB
Too big Get file →
Q3_K_M467.3 GB (13 parts) 468.9 GB
Too big Get file →
Q3_K_L487.5 GB (13 parts) 489.1 GB
Too big Get file →
Q3_K_XL488.6 GB (13 parts) 490.2 GB
Too big Get file →
IQ4_XS547.6 GB (14 parts) 549.2 GB
Too big Get file →
IQ4_NL579.5 GB (16 parts) 581.1 GB
Too big Get file →
Q4_K_S579.8 GB (16 parts) 581.4 GB
Too big Get file →
Q4_K_M580.4 GB (16 parts) 581.9 GB
Too big Get file →
Q5_K_S580.5 GB (16 parts) 582.1 GB
Too big Get file →
Q4_K_L581.2 GB (16 parts) 582.8 GB
Too big Get file →
Q5_K_M581.2 GB (16 parts) 582.8 GB
Too big Get file →
Q6_K582.2 GB (16 parts) 583.8 GB
Too big Get file →
Q8_0583.7 GB (16 parts) 585.3 GB
Too big Get file →
Q4_0589.8 GB (16 parts) 591.4 GB
Too big Get file →
Q4_1643.2 GB (17 parts) 644.7 GB
Too big Get file →

For 8 GB VRAM, use Q4_K_M (file size 580.36 GB, memory 581.9 GB at 8k context). For 12 GB VRAM, Q5_K_M (580.36 GB, 582.8 GB). For 16 GB VRAM, Q8_0 (583.71 GB, 585.3 GB).

What GPU do you need for Kimi K2.5?

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 (1.5 GB file, about 3.1 GB in use).
  • 6 GB VRAM: the best quant that fits is the main file (1.5 GB file, about 3.1 GB in use).
  • 8 GB VRAM: the best quant that fits is the main file (1.5 GB file, about 3.1 GB in use).
  • 12 GB VRAM: the best quant that fits is the main file (1.5 GB file, about 3.1 GB in use).
  • 16 GB VRAM: the best quant that fits is the main file (1.5 GB file, about 3.1 GB in use).
  • 24 GB VRAM: the best quant that fits is the main file (1.5 GB file, about 3.1 GB in use).
  • 32 GB VRAM: the best quant that fits is the main file (1.5 GB file, about 3.1 GB in use).
  • 48 GB VRAM: the best quant that fits is the main file (1.5 GB file, about 3.1 GB in use).

Which Kimi K2.5 quant should you choose?

Q4_K_M is the usual starting point for Kimi K2.5: it balances file size and quality well. Here it is a 580.4 GB file.

The Q8_0 file is 583.7 GB, about 1.0 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 1.5 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

Architecturedeepseek2
Parameters1030B (approximate, mixture of experts)
Layers61
Trained context262,144 tokens
KV cache (fp16)130 KB per token

Source

Quantized bybartowski
FamilyKimi
Licenseother
Downloads (HF)2,091
Last updated2026-02-17

Frequently asked questions about Kimi K2.5

How much VRAM does Kimi K2.5 need?

About 581.9 GB at Q4_K_M with an 8K context window. That is the 580.4 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 Kimi K2.5 on an 8 GB GPU?

Yes, with the smaller files: the 1.5 GB file needs about 3.1 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Kimi K2.5 GGUF?

It depends on the quant: file is 1.5 GB, IQ1_S is 210.3 GB, IQ1_M is 219.6 GB, IQ2_XXS is 245.3 GB, IQ2_XS is 283.1 GB, IQ2_S is 283.8 GB and 21 more.

What context length does Kimi K2.5 support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.3 GB at 32K.

Where can I download Kimi K2.5 GGUF?

The files are on Hugging Face in the bartowski/moonshotai_Kimi-K2.5-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 Kimi K2.5 use?

Hugging Face lists the license as other. 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 Kimi K2.5 on 8 GB VRAM?

Use the Q4_K_M quant; it requires about 581.9 GB of VRAM at 8k context, which is not feasible on 8 GB. Smaller quants are not available.

Does Kimi K2.5 support vision?

Yes, the model has native multimodality and can process visual inputs.

What is the context length of Kimi K2.5?

The trained context length is 262,144 tokens.

Other GGUFs of Kimi-K2.5

Similar models

Browse all →

More from bartowski

All by bartowski →

More Kimi models

All Kimi →

Other large models

All large models →