Model

Qwen3-4B-Thinking-2507

qwen3-4b-thinking-2507

JSON API Find compatible recipes

Record

Family: qwen
Architecture: dense
Parameters: 4
HF downloads / 30d: 20,200

Hugging Face model card

Identity

https://huggingface.co/lmstudio-community/Qwen3-4B-Thinking-2507-GGUF

Repository: lmstudio-community/Qwen3-4B-Thinking-2507-GGUF
Status: known
Link type: Exact Hub repository

Public Hugging Face repository confirmed by the Hub API.

Leaderboard scores

Scraped public quality scores joined to this model by its Hugging Face repository. They describe the listed variant, not any specific local quantization, and are never speed measurements.

Benchmark	Category	Score	Rank	Scored variant	Confidence
bfcl	agentic	69.1	11	Qwen3-4B-Thinking-2507	medium
bfcl	agentic	69.1	12	Qwen3-4B-Thinking-2507-FP8	medium
bfcl	agentic	69.1	13	Qwen3-4B-Thinking-2507-unsloth-bnb-4bit	medium
bfcl	agentic	69.1	14	Qwen3-4B-Thinking-2507-GGUF	medium
bfcl	agentic	69.1	15	Qwen3-4B-Thinking-2507	medium
bfcl	agentic	69.1	16	Qwen3-4B-Thinking-2507-bnb-4bit	medium
bfcl	agentic	69.1	18	Qwen3-4B-Thinking-2507-GGUF	medium
chatbot-arena	chat	36.3	135	Qwen3-4B-Thinking-2507	medium
chatbot-arena	chat	36.3	136	Qwen3-4B-Thinking-2507-FP8	medium
chatbot-arena	chat	36.3	137	Qwen3-4B-Thinking-2507-unsloth-bnb-4bit	medium
chatbot-arena	chat	36.3	138	Qwen3-4B-Thinking-2507-GGUF	medium
chatbot-arena	chat	36.3	139	Qwen3-4B-Thinking-2507	medium
chatbot-arena	chat	36.3	140	Qwen3-4B-Thinking-2507-bnb-4bit	medium
chatbot-arena	chat	36.3	142	Qwen3-4B-Thinking-2507-GGUF	medium
livecodebench	coding	57.4	257	Qwen3-4B-Thinking-2507	medium
livecodebench	coding	57.4	258	Qwen3-4B-Thinking-2507-FP8	medium
livecodebench	coding	57.4	259	Qwen3-4B-Thinking-2507-unsloth-bnb-4bit	medium
livecodebench	coding	57.4	260	Qwen3-4B-Thinking-2507-GGUF	medium
livecodebench	coding	57.4	261	Qwen3-4B-Thinking-2507	medium
livecodebench	coding	57.4	262	Qwen3-4B-Thinking-2507-bnb-4bit	medium
livecodebench	coding	57.4	264	Qwen3-4B-Thinking-2507-GGUF	medium
if-eval	instruction-following	86.5	58	Qwen3-4B-Thinking-2507	medium
if-eval	instruction-following	86.5	59	Qwen3-4B-Thinking-2507-FP8	medium
if-eval	instruction-following	86.5	60	Qwen3-4B-Thinking-2507-unsloth-bnb-4bit	medium
if-eval	instruction-following	86.5	61	Qwen3-4B-Thinking-2507-GGUF	medium
if-eval	instruction-following	86.5	62	Qwen3-4B-Thinking-2507	medium
if-eval	instruction-following	86.5	63	Qwen3-4B-Thinking-2507-bnb-4bit	medium
if-eval	instruction-following	86.5	65	Qwen3-4B-Thinking-2507-GGUF	medium
mmlu	knowledge	89.5	29	Qwen3-4B-Thinking-2507	medium
mmlu	knowledge	89.5	30	Qwen3-4B-Thinking-2507-FP8	medium
mmlu	knowledge	89.5	32	Qwen3-4B-Thinking-2507-unsloth-bnb-4bit	medium
mmlu	knowledge	89.5	33	Qwen3-4B-Thinking-2507-GGUF	medium
mmlu	knowledge	89.5	34	Qwen3-4B-Thinking-2507	medium
mmlu	knowledge	89.5	35	Qwen3-4B-Thinking-2507-bnb-4bit	medium
mmlu	knowledge	89.5	37	Qwen3-4B-Thinking-2507-GGUF	medium
mmlu-pro	knowledge	78.5	175	Qwen3-4B-Thinking-2507	medium
mmlu-pro	knowledge	78.5	176	Qwen3-4B-Thinking-2507-FP8	medium
mmlu-pro	knowledge	78.5	177	Qwen3-4B-Thinking-2507-unsloth-bnb-4bit	medium
mmlu-pro	knowledge	78.5	178	Qwen3-4B-Thinking-2507-GGUF	medium
mmlu-pro	knowledge	78.5	179	Qwen3-4B-Thinking-2507	medium
mmlu-pro	knowledge	78.5	180	Qwen3-4B-Thinking-2507-bnb-4bit	medium
mmlu-pro	knowledge	78.5	182	Qwen3-4B-Thinking-2507-GGUF	medium
mmlu-pro	knowledge	73.1	229	Qwen3-4B-Thinking-2507	medium
mmlu-pro	knowledge	73.1	230	Qwen3-4B-Thinking-2507-FP8	medium
mmlu-pro	knowledge	73.1	231	Qwen3-4B-Thinking-2507-unsloth-bnb-4bit	medium
mmlu-pro	knowledge	73.1	232	Qwen3-4B-Thinking-2507-GGUF	medium
mmlu-pro	knowledge	73.1	233	Qwen3-4B-Thinking-2507	medium
mmlu-pro	knowledge	73.1	234	Qwen3-4B-Thinking-2507-bnb-4bit	medium
mmlu-pro	knowledge	73.1	236	Qwen3-4B-Thinking-2507-GGUF	medium
aime-2025	math	70.9	130	Qwen3-4B-Thinking-2507	medium
aime-2025	math	70.9	131	Qwen3-4B-Thinking-2507-FP8	medium
aime-2025	math	70.9	132	Qwen3-4B-Thinking-2507-unsloth-bnb-4bit	medium
aime-2025	math	70.9	133	Qwen3-4B-Thinking-2507-GGUF	medium
aime-2025	math	70.9	134	Qwen3-4B-Thinking-2507	medium
aime-2025	math	70.9	135	Qwen3-4B-Thinking-2507-bnb-4bit	medium
aime-2025	math	70.9	137	Qwen3-4B-Thinking-2507-GGUF	medium
gpqa	reasoning	65.8	85	Qwen3-4B-Thinking-2507	medium
gpqa	reasoning	65.8	86	Qwen3-4B-Thinking-2507-FP8	medium
gpqa	reasoning	65.8	91	Qwen3-4B-Thinking-2507-unsloth-bnb-4bit	medium
gpqa	reasoning	65.8	92	Qwen3-4B-Thinking-2507-GGUF	medium
gpqa	reasoning	65.8	94	Qwen3-4B-Thinking-2507	medium
gpqa	reasoning	65.8	95	Qwen3-4B-Thinking-2507-bnb-4bit	medium
gpqa	reasoning	65.8	105	Qwen3-4B-Thinking-2507-GGUF	medium

Remaining fields

Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.

architecture: dense
family: qwen
id: qwen3-4b-thinking-2507
name: Qwen3-4B-Thinking-2507
params: 4
schema version: local-ai-registry/v1
url: huggingface.co/lmstudio-community/Qwen3-4B-Thinking-2507-GGUF ↗

downloads

all time: 314,606
captured at: 2026-08-31T12:36:33Z
last 30d: 20,200

Provenance & metadata (1)

source huggingface-api

Provenance & metadata (2)

facts

active params · provenance · captured at: 2026-08-30T09:10:02Z

active params · reason not-observed

active params · state unknown

provenance

captured at: 2026-08-30T09:10:02Z

sources

captured at	kind	url
2026-08-30T09:10:02Z	normalized-model	huggingface.co/lmstudio-community/Qwen3-4B-Thinking-2507-GGUF ↗