Qwen3 VL 30B A3B Reasoning vs Qwen3 VL 4B (Reasoning)
Wins 6 of 7 areas
Coding · Reasoning · Facts · Images and charts · Long documents · Following instructions
Wins 1 of 7 areas
Agents
Qwen3 VL 30B A3B Reasoning is the stronger all-rounder.Qwen3 VL 4B (Reasoning) is better at agents.
Scores updated · 52 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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.
- Images and chartsUnderstanding pictures, charts and video70Qwen3 VL 30B A3B Reasoning7 of 7 tests
- CodingWriting and fixing software30Qwen3 VL 30B A3B Reasoning3 of 3 tests
- ReasoningHard problems that need careful thinking20Qwen3 VL 30B A3B Reasoning2 of 3 tests · 1 tie
- FactsGetting facts right instead of making them up20Qwen3 VL 30B A3B Reasoning2 of 2 tests
- Long documentsFinding answers in very long texts10Qwen3 VL 30B A3B Reasoning1 of 1 test
- Following instructionsDoing exactly what it is asked10Qwen3 VL 30B A3B Reasoning1 of 1 test
- AgentsCarrying out multi-step tasks on its own01Qwen3 VL 4B (Reasoning)1 of 2 tests · 1 tie
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 tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Qwen3 VL 30B A3B Reasoning pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+22.6points ahead
- Reasons across sets of long documentsAA-LCR+21.4points ahead
- Questions about short and long videosVideo-MME+13.6points ahead
Where Qwen3 VL 4B (Reasoning) pulls ahead
- Real work tasks from 44 professionsGDPVal+5.5points ahead
Every test, side by side
All 52 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingQwen3 VL 30B A3B Reasoning
- LiveCodeBench v6Qwen3 VL 30B A3B Reasoning by 12.964.251.3+12.9
- SciCodeQwen3 VL 30B A3B Reasoning by 11.728.817.1+11.7
- Terminal-Bench HardQwen3 VL 30B A3B Reasoning by 3.85.31.5+3.8
AgentsQwen3 VL 4B (Reasoning)
- GDPValQwen3 VL 4B (Reasoning) by 5.58.313.8+5.5
- OSWorld-Verifiedtie30.631.4tie
ReasoningQwen3 VL 30B A3B Reasoning
- GPQA DiamondQwen3 VL 30B A3B Reasoning by 22.67249.4+22.6
- Humanity's Last ExamQwen3 VL 30B A3B Reasoning by 4.38.94.6+4.3
- CritPttie00tie
FactsQwen3 VL 30B A3B Reasoning
- AA-Omniscience · AccuracyQwen3 VL 30B A3B Reasoning by 4.716.912.2+4.7
- AA-Omniscience · Non-hallucinationQwen3 VL 30B A3B Reasoning by 210.28.2+2
Images and chartsQwen3 VL 30B A3B Reasoning
- Video-MMEQwen3 VL 30B A3B Reasoning by 13.673.359.7+13.6
- MMMU-ProQwen3 VL 30B A3B Reasoning by 9.861.852+9.8
- CharXiv (RQ)Qwen3 VL 30B A3B Reasoning by 6.356.650.3+6.3
- OCRBenchQwen3 VL 30B A3B Reasoning by 3.183.980.8+3.1
- MathVistaQwen3 VL 30B A3B Reasoning by 2.481.979.5+2.4
- MMStarQwen3 VL 30B A3B Reasoning by 2.375.573.2+2.3
- BLINKQwen3 VL 30B A3B Reasoning by 265.463.4+2
Long documentsQwen3 VL 30B A3B Reasoning
- AA-LCRQwen3 VL 30B A3B Reasoning by 21.442.721.3+21.4
Following instructionsQwen3 VL 30B A3B Reasoning
- IFBenchQwen3 VL 30B A3B Reasoning by 8.545.136.6+8.5
Other results33 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.
- Arena-Hard-V2Qwen3 VL 30B A3B Reasoning by 19.956.736.8+19.9
- HMMT 2025Qwen3 VL 30B A3B Reasoning by 14.567.653.1+14.5
- MMLU-ProXQwen3 VL 30B A3B Reasoning by 11.176.165+11.1
- AA-OmniscienceQwen3 VL 30B A3B Reasoning by 10.8-57.6-68.4+10.8
- IncludeQwen3 VL 30B A3B Reasoning by 9.974.564.6+9.9
- SuperGPQAQwen3 VL 30B A3B Reasoning by 9.656.446.8+9.6
- AIME 2025Qwen3 VL 30B A3B Reasoning by 8.683.174.5+8.6
- ScreenSpot-Pro (No tools)Qwen3 VL 30B A3B Reasoning by 8.157.349.2+8.1
- MMLU-ProQwen3 VL 30B A3B Reasoning by 6.980.573.6+6.9
- Artificial Analysis Coding IndexQwen3 VL 30B A3B Reasoning by 6.413.16.7+6.4
- LVBenchQwen3 VL 30B A3B Reasoning by 5.759.253.5+5.7
- MATH-VisionQwen3 VL 30B A3B Reasoning by 5.765.760+5.7
- VideoMMMUQwen3 VL 30B A3B Reasoning by 5.67569.4+5.6
- MMMU (val) (Pass@1)Qwen3 VL 30B A3B Reasoning by 5.27670.8+5.2
- MMLU-ReduxQwen3 VL 30B A3B Reasoning by 4.990.986+4.9
- OCRBench v2 (Chinese)Qwen3 VL 30B A3B Reasoning by 4.660.455.8+4.6
- τ²-Bench Telecom (AA run)Qwen3 VL 30B A3B Reasoning by 4.419.915.5+4.4
- RealWorldQAQwen3 VL 30B A3B Reasoning by 4.277.473.2+4.2
- CC-OCRQwen3 VL 30B A3B Reasoning by 477.873.8+4
- CharadesSTAQwen3 VL 30B A3B Reasoning by 3.762.759+3.7
- InfoVQA_testQwen3 VL 30B A3B Reasoning by 38683+3
- MV-BenchQwen3 VL 30B A3B Reasoning by 2.77269.3+2.7
- AA IntelligenceQwen3 VL 30B A3B Reasoning by 2.59.57+2.5
- AI2DQwen3 VL 30B A3B Reasoning by 286.984.9+2
- ERQAQwen3 VL 4B (Reasoning) by 245.347.3+2
- HallusionBenchQwen3 VL 30B A3B Reasoning by 1.96664.1+1.9
- ScreenSpotQwen3 VL 30B A3B Reasoning by 1.894.792.9+1.8
- BFCL v3Qwen3 VL 30B A3B Reasoning by 1.368.667.3+1.3
- IFEvaltie81.782.6tie
- DocVQA_testtie9594.2tie
- OCRBench v2_entie62.661.8tie
- AA Agentic Indextie14.914.4tie
- MM MTBenchtie7.97.7tie
Questions people ask
Which is better, Qwen3 VL 30B A3B Reasoning or Qwen3 VL 4B (Reasoning)?
Qwen3 VL 30B A3B Reasoning wins six of the seven areas where both have results: coding, reasoning, facts, images and charts, long documents and following instructions. Qwen3 VL 4B (Reasoning) wins agents.
Which is better for coding?
Qwen3 VL 30B A3B Reasoning. It wins 3 of the 3 coding tests both models report; Qwen3 VL 4B (Reasoning) wins none.
How do you compare the two?
We use the 52 benchmark tests both models have published scores on. The verdict counts the 19 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 33 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.