Qwen3.5 4B vs Qwen3 VL 4B Instruct
Wins 5 of 6 areas
Coding · Reasoning · Facts · Long documents · Following instructions
Wins 1 of 6 areas
Images and charts
Qwen3.5 4B is the stronger all-rounder.Qwen3 VL 4B Instruct is better at images and charts.
Scores updated · 35 tests both models report · How we compare
Where each one wins
Tests won in each of the six areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software30Qwen3.5 4B3 of 3 tests
- ReasoningHard problems that need careful thinking20Qwen3.5 4B2 of 3 tests · 1 tie
- FactsGetting facts right instead of making them up20Qwen3.5 4B2 of 2 tests
- Long documentsFinding answers in very long texts10Qwen3.5 4B1 of 1 test
- Following instructionsDoing exactly what it is asked10Qwen3.5 4B1 of 1 test
- Images and chartsUnderstanding pictures, charts and video14Qwen3 VL 4B Instruct4 of 5 tests
Long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Qwen3.5 4B pulls ahead
- Reasons across sets of long documentsAA-LCR+49points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+40points ahead
- Harder college exam questions with imagesMMMU-Pro+21.5points ahead
Every test, side by side
All 35 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingQwen3.5 4B
- Terminal-Bench HardQwen3.5 4B by 18.218.20+18.2
- LiveCodeBench v6Qwen3.5 4B by 17.955.837.9+17.9
- SciCodeQwen3.5 4B by 2.416.113.7+2.4
ReasoningQwen3.5 4B
- GPQA DiamondQwen3.5 4B by 4077.137.1+40
- Humanity's Last ExamQwen3.5 4B by 6.39.93.6+6.3
- CritPttie00tie
FactsQwen3.5 4B
- AA-Omniscience · Non-hallucinationQwen3.5 4B by 10.813.42.6+10.8
- AA-Omniscience · AccuracyQwen3.5 4B by 4.215.110.9+4.2
Images and chartsQwen3 VL 4B Instruct
Following instructionsQwen3.5 4B
- IFBenchQwen3.5 4B by 20.25231.8+20.2
Other results20 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- τ²-Bench Telecom (AA run)Qwen3.5 4B by 68.792.123.4+68.7
- HMMT 2025Qwen3.5 4B by 46.176.830.7+46.1
- AA Agentic IndexQwen3.5 4B by 24.732.57.8+24.7
- Artificial Analysis Coding IndexQwen3.5 4B by 1822.64.6+18
- AA-OmniscienceQwen3.5 4B by 17.7-58.4-76+17.7
- MMMU (val) (Pass@1)Qwen3 VL 4B Instruct by 17.150.367.4+17.1
- SuperGPQAQwen3.5 4B by 12.652.940.3+12.6
- MMLU-ProXQwen3.5 4B by 12.171.559.4+12.1
- MMLU-ProQwen3.5 4B by 1279.167.1+12
- IncludeQwen3.5 4B by 9.67161.4+9.6
- BFCL v3Qwen3.5 4B by 7.871.163.3+7.8
- IFEvalQwen3.5 4B by 7.589.882.3+7.5
- AA IntelligenceQwen3.5 4B by 7.413.15.7+7.4
- MMLU-ReduxQwen3.5 4B by 7.388.881.5+7.3
- HallusionBenchQwen3 VL 4B Instruct by 5.951.757.6+5.9
- ERQAQwen3.5 4B by 546.341.3+5
- OCRBench v2_enQwen3 VL 4B Instruct by 558.763.7+5
- RealWorldQAQwen3 VL 4B Instruct by 3.867.170.9+3.8
- TextVQA-valtie81.280.5tie
- DocVQA-valtie94.894.7tie
Questions people ask
Which is better, Qwen3.5 4B or Qwen3 VL 4B Instruct?
Qwen3.5 4B wins five of the six areas where both have results: coding, reasoning, facts, long documents and following instructions. Qwen3 VL 4B Instruct wins images and charts.
Which is better for coding?
Qwen3.5 4B. It wins 3 of the 3 coding tests both models report; Qwen3 VL 4B Instruct wins none.
How do you compare the two?
We use the 35 benchmark tests both models have published scores on. The verdict counts the 15 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 20 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.