qwen3.5-35b-a3b-t48-c8k:latest

qwen35moe · 36.0B · Q4_K_M

THINKING MODEL

Dell Inc. PowerEdge R820 (Intel Xeon® E5-4620 0)

256 GB · Microsoft Windows Server 2022 Datacenter 10.0.20348

Tested on July 30, 2026 · Submitted by Csaba_R820

Top 100% Compare

Global Score

10 /100

Not Rec.

Hardware Fit

33/100

Quality

0/100

Get this model

🦙

Ollama

ollama pull qwen3.5-35b-a3b-t48-c8k:latest

View on Ollama Library

ollama.com/library/qwen3.5-35b-a3b-t48-c8k

Get it in LM Studio

Search and download models directly from the app

🤗

Find on HuggingFace

GGUF versions & quantizations

Hardware

Machine: Dell Inc. PowerEdge R820
CPU: Intel Xeon® E5-4620 0
Cores: 64 total (64 perf)
Frequency: 2.2 GHz
RAM: 256 GB DDR3
GPU: Microsoft Remote Display Adapter, Microsoft Basic Display Adapter
OS: Microsoft Windows Server 2022 Datacenter 10.0.20348
Arch: x64
Power Mode: balanced

Performance

Tokens/sec Tokens generated per second — higher is better: 3.5
Standard deviation Variation of speed across benchmark runs — lower means more consistent: ±0.0
First chunk latency Delay before the first streamed chunk arrives from runtime: 3.5 s
Time to first token How fast the model starts responding — lower is better: 30.0 s
Load time Time to load the model into memory before inference: 30.0 s
Memory usage RAM consumed during inference — percentage of total system memory: 22.0 GB (9%)
Total tokens Total number of tokens generated across all benchmark prompts: 1429
Thinking tokens (est.) Estimated internal reasoning tokens — models with "thinking" generate more tokens: ~789

Score breakdown

Speed

3/50

Time to first token

0/20

Memory

30/30

Quality

Reasoning

0/20

Coding

0/20

Instruction following

0/20

Structured output

0/15

Math

0/15

Multilingual

0/10

Category levels

Reasoning: Poor Coding: Poor Instruction Following: Poor Structured Output: Poor Math: Poor Multilingual: Poor

Metadata

Spec version: 0.2.1
Runtime: Ollama 0.32.4
Model format: GGUF
Hardware profile: HIGH-END
Result hash: eef949d7c736cadb9762f0fd3c9e6da46078cb1608733545167019360ec307f7

Interpretation

Hardware fit: 33/100. Overall suitability: NOT RECOMMENDED (Global 10/100). Category profile: Reasoning: Poor, Coding: Poor, Instruction Following: Poor, Structured Output: Poor, Math: Poor, Multilingual: Poor. Warning: model produced very low accuracy on quality tasks — results may be unusable despite good hardware performance.

Warnings

Model produced very low accuracy on quality tasks — results may be unusable despite good hardware performance.

Disqualifiers

Token speed too low: 3.5 tok/s (minimum: 8 tok/s for HIGH-END profile)
Time to first token too high: 30000ms (maximum: 10000ms for HIGH-END profile)

Bench Environment

CPU load: avg 61% (peak 74%)

Run yours now

$ npm install -g metrillm@latest

$ metrillm

Requires Node 20+ and Ollama or LM Studio running

Or run without installing: npx metrillm@latest