Qwen3-VL-32B-Instruct against Qwen3-VL-8B-Instruct

Pick two models to see them side by side: Overall score, every category, the questions where one did better, and speed and memory. Each model is shown at its newest tested version. Back to the leaderboard.

The short version

Same model line in two sizes: Qwen3-VL-8B-Instruct has 8.8 billion parameters and Qwen3-VL-32B-Instruct has 33.4 billion parameters.

Qwen3-VL-8B-Instruct scores higher overall: 94.9 against 94.7 out of 100, a gap of 0.2 points.

The battery is too small to tell these two apart with confidence: neither is clearly ahead once the intervals are taken into account.

Across the 9 categories that count toward Overall, Qwen3-VL-32B-Instruct is higher in 1, Qwen3-VL-8B-Instruct is higher in 4 and 4 are level.

The widest gap is in Coding: 86.5 for Qwen3-VL-32B-Instruct against 77.7 for Qwen3-VL-8B-Instruct, in favour of Qwen3-VL-32B-Instruct.

Of the 259 questions both models have a definite score for, Qwen3-VL-32B-Instruct did better on 12, Qwen3-VL-8B-Instruct did better on 10, and 237 were level.

The two were tested in different runs, so their hard-question scores are not compared.

Their speeds were not measured on one shared backend and Ollama version, so they are not compared.

Written from the numbers below. Scores come from one fixed battery run on one laptop; read the methodology before drawing conclusions.

Overall

Qwen3-VL-8B-Instruct scores higher: 94.9 against 94.7, a gap of 0.2 points. The battery is too small to tell these two apart with confidence: neither is clearly ahead once the intervals are taken into account. How sure is a score?

Overall (0–100) with its 95% interval, the rank range among all listed ranked models, and the hard-question score.
MeasureQwen3-VL-32B-InstructQwen3-VL-8B-Instruct
Overall94.792.1–96.994.992.4–97.1
Rank range4–604–48
Hard questionsmean score on the questions that half or fewer of the ranked models of 3B and over in the model's own run answered correctly51.066.1
Size class16–35B4–9B

By category

Programmatic mean (0–100) per category. Higher is the model with the larger mean; a gap that rounds to 0.0 is called level. A dash means not scored, which is different from 0.
CategoryQwen3-VL-32B-InstructQwen3-VL-8B-InstructHigherDifference
Reasoningcounts toward Overall 88.6 91.4 Qwen3-VL-8B-Instruct 2.8
Mathcounts toward Overall 88.0 92.0 Qwen3-VL-8B-Instruct 4.0
Codingcounts toward Overall 86.5 77.7 Qwen3-VL-32B-Instruct 8.8
Instruction followingcounts toward Overall 96.1 98.5 Qwen3-VL-8B-Instruct 2.4
Knowledge & honestycounts toward Overall 100.0 100.0 Level 0
Long contextcounts toward Overall 94.4 94.4 Level 0
Multilingualcounts toward Overall 98.3 100.0 Qwen3-VL-8B-Instruct 1.7
Tool usecounts toward Overall 100.0 100.0 Level 0
Safetycounts toward Overall 100.0 100.0 Level 0
Pushback resistanceshown separately 100.0 100.0 Level 0
Famous gotchasshown separately 93.3 93.3 Level 0
Writingshown separately — — — —
Self-descriptionshown separately — — — —

Speed and memory

Measured on one laptop. Decode speed is generation speed with one request at a time.
MeasureQwen3-VL-32B-InstructQwen3-VL-8B-Instruct
Decode speed, one stream (tok/s)3.29.4
GPU memory on our server (1 to 8 answers at a time) (GB)24.57.8
Model size as loaded on our server (GB)24.57.8
BackendVulkan (Ollama 0.34.4, flash attention, 16k context; 4 parallel slots)ROCm 7.1 (Ollama 0.34.4, flash attention, 16k context; 4 parallel slots until 2026-10-07 00:45Z, 8 after)

A dash means the figure was not measured, which is different from 0. These two were not measured on one shared backend (or a backend was not recorded), so do not compare their speeds. Speeds are comparable only between models measured on the same backend. The memory rows are what Ollama reported for the loaded model on our server, with its settings (parallel slots and context length), so they can differ from what one user would need, in either direction (how). They are not the memory estimate the hardware picker uses. See also the Speed page.

Where they differ

Both models have a definite score for 259 questions. Qwen3-VL-32B-Instruct did better on 12, Qwen3-VL-8B-Instruct did better on 10, and 237 were level. A score gap of more than 0.001 counts as a win. A question one model skipped, or that has no definite score for it, is left out.

Questions both models have a score for, by category.
CategoryQuestionsQwen3-VL-32B-Instruct betterQwen3-VL-8B-Instruct betterLevel
Reasoning350134
Math250124
Coding3010218
Instruction following301425
Knowledge & honesty300030
Long context181116
Multilingual200119
Tool use200020
Safety240024
Pushback resistance120012
Famous gotchas150015

Questions where only one got full marks

Full marks means a score of 1. The answers of public questions can be opened below. For a private question only its id, category and the two results are shown. Each result links to that question on the model's page.

Reasoning

Reasoning: 1 question where exactly one model got full marks.
QuestionQwen3-VL-32B-InstructQwen3-VL-8B-Instruct
reasoning-025private fail on its page pass on its page

Math

Math: 1 question where exactly one model got full marks.
QuestionQwen3-VL-32B-InstructQwen3-VL-8B-Instruct
math-022private fail on its page pass on its page

Coding

Coding: 5 questions where exactly one model got full marks.
QuestionQwen3-VL-32B-InstructQwen3-VL-8B-Instruct
coding-011private pass on its page fail on its page
coding-021private pass on its page 89% on its page
coding-023private pass on its page 78% on its page
coding-025private pass on its page 89% on its page
coding-026private pass on its page fail on its page

Instruction following

Instruction following: 4 questions where exactly one model got full marks.
QuestionQwen3-VL-32B-InstructQwen3-VL-8B-Instruct
instruction-005private 67% on its page pass on its page
instruction-012private pass on its page 75% on its page
instruction-020private 80% on its page pass on its page
instruction-021private 75% on its page pass on its page

Long context

Long context: 2 questions where exactly one model got full marks.
QuestionQwen3-VL-32B-InstructQwen3-VL-8B-Instruct
long_context-005private pass on its page fail on its page
long_context-014private fail on its page pass on its page

Multilingual

Multilingual: 1 question where exactly one model got full marks.
QuestionQwen3-VL-32B-InstructQwen3-VL-8B-Instruct
multilingual-020private 67% on its page pass on its page

The same numbers as JSON: compare.json. Answers appear there only for public questions.