GGUF models by Thireus

2513 GGUF models quantized by Thireus on Hugging Face. Sizes run from 275M to 5632B parameters. The longest trained context is 1024K tokens.

Most downloadedNewestLargestLongest contextSmallest
GL
GLM 4.6 THIREUS BF16 SPECIAL SPLIT
GLM · Thireus
Params357B
Context198K
Smallest713.7 GB
ReasoningTool use
GL
GLM 4.5 Air
GLM · Thireus
Params110B
Context128K
Smallest22.6 GB
ReasoningTool use
GL
GLM 4.5 Air
GLM · Thireus
Params1203.2B
Context128K
Smallest27.9 GB
GGUF
QW
Qwen3 235B A22B Thinking 2507
Qwen · Thireus
Params235B
Context256K
Smallest194.9 GB
ReasoningTool use
QW
Qwen3.6 35B A3B
Qwen · Thireus
Params34.7B
Context256K
Smallest6.7 GB
ReasoningTool use
KI
Kimi K2 Instruct 0905 THIREUS Q8 0 SPECIAL SPLIT
Kimi · Thireus
Params1030B
Context256K
Smallest1091.1 GB
GGUF
QW
Qwen3 Coder 480B A35B Instruct
Qwen · Thireus
Params480B
Context256K
Smallest255.8 GB
CodingTool use
QW
Qwen3 4B Instruct 2507
Qwen · Thireus
Params4.02B
Context256K
Smallest3.3 GB
ReasoningTool use
GL
GLM 4.5 Air
GLM · Thireus
Params120B
Context128K
Smallest44.4 GB
GGUF
GL
GLM 4.5 Air
GLM · Thireus
Params1203.2B
Context128K
Smallest57.5 GB
GGUF
DE
DeepSeek V3 0324
DeepSeek · Thireus
Params672B
Context160K
Smallest442.1 GB
Tool use
KI
Kimi K2 Instruct 0905 THIREUS Q5 K SPECIAL SPLIT
Kimi · Thireus
Params1030B
Context256K
Smallest706.2 GB
GGUF
QW
Qwen3 235B A22B Thinking 2507
Qwen · Thireus
Params235B
Context256K
Smallest59.3 GB
ReasoningTool use
DE
DeepSeek R1 0528
DeepSeek · Thireus
Params672B
Context160K
Smallest358.1 GB
ReasoningTool use
GL
GLM 4.5 Air
GLM · Thireus
Params1203.2B
Context128K
Smallest80.3 GB
GGUF
GL
GLM 4.5 THIREUS BF16 SPECIAL SPLIT
GLM · Thireus
Params360B
Context128K
Smallest716.8 GB
GGUF
QW
Qwen3 Coder 480B A35B Instruct
Qwen · Thireus
Params480B
Context256K
Smallest142.8 GB
CodingTool use
GL
GLM 4.5 Air
GLM · Thireus
Params1203.2B
Context128K
Smallest42.0 GB
GGUF
QW
Qwen3 235B A22B Thinking 2507
Qwen · Thireus
Params235B
Context256K
Smallest147.1 GB
ReasoningTool use
GL
GLM 4.5 Air
GLM · Thireus
Params120B
Context128K
Smallest44.4 GB
GGUF
GL
GLM 4.5 Air
GLM · Thireus
Params1203.2B
Context128K
Smallest113.3 GB
GGUF
GL
GLM 4.5 Air
GLM · Thireus
Params1203.2B
Context128K
Smallest73.7 GB
GGUF
QW
Qwen3 235B A22B Thinking 2507
Qwen · Thireus
Params235B
Context256K
Smallest70.0 GB
ReasoningTool use
KI
Kimi K2 Instruct 0905 THIREUS IQ3 S R4 SPECIAL SPLIT
Kimi · Thireus
Params1000B
Context256K
Smallest441.7 GB
GGUF
GL
GLM 4.5 Air
GLM · Thireus
Params1203.2B
Context128K
Smallest52.1 GB
GGUF
QW
Qwen3 4B Thinking 2507
Qwen · Thireus
Params4.02B
Context256K
Smallest2.1 GB
ReasoningTool use
GL
GLM 4.5 Air
GLM · Thireus
Params110B
Context128K
Smallest57.5 GB
ReasoningTool use
QW
Qwen3 4B Instruct 2507
Qwen · Thireus
Params4.02B
Context256K
Smallest2.0 GB
ReasoningTool use
DE
DeepSeek R1 0528
DeepSeek · Thireus
Params672B
Context160K
Smallest378.4 GB
ReasoningTool use
QW
Qwen3 235B A22B Thinking 2507
Qwen · Thireus
Params235B
Context256K
Smallest193.0 GB
ReasoningTool use

About Thireus on Hugging Face

Thireus has published 2,513 GGUF models that ggufmodels.com tracks, across families such as Qwen, GLM and Kimi. The most downloaded are GLM 4.6 THIREUS BF16 SPECIAL SPLIT, GLM 4.5 Air and GLM 4.5 Air.

The quants Thireus publishes most often are Q8_0 (82), Q6_K (82), Q6_0 (82) and Q5_0 (82). Parameter counts run from 275M to 5632B, and the longest trained context window is 1024K tokens.

Frequently asked questions

What GGUF models has Thireus published?

2,513 models are listed here, mainly from the Qwen, GLM and Kimi families. The most downloaded is GLM 4.6 THIREUS BF16 SPECIAL SPLIT.

Which quants does Thireus publish?

The most common are Q8_0 (82 models), Q6_K (82 models), Q6_0 (82 models) and Q5_0 (82 models).

Keep browsing