Gemini 2.5 Flash vs Qwen3.5 27B
Wins 1 of 7 areas
Facts
Wins 6 of 7 areas
Coding · Agents · Reasoning · Images and charts · Long documents · Following instructions
Qwen3.5 27B is the stronger all-rounder.Gemini 2.5 Flash is better at facts.
Scores updated · 24 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.
- CodingWriting and fixing software03Qwen3.5 27B3 of 3 tests
- ReasoningHard problems that need careful thinking02Qwen3.5 27B2 of 3 tests · 1 tie
- Images and chartsUnderstanding pictures, charts and video02Qwen3.5 27B2 of 3 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own01Qwen3.5 27B1 of 1 test
- Long documentsFinding answers in very long texts01Qwen3.5 27B1 of 1 test
- Following instructionsDoing exactly what it is asked01Qwen3.5 27B1 of 1 test
- FactsGetting facts right instead of making them up30Gemini 2.5 Flash3 of 3 tests
Agents, 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.
Qwen3.5 27B costs 4% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Qwen3.5 27B pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+25.3points ahead
- Real work tasks from 44 professionsGDPVal+23.1points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+19points ahead
Where Gemini 2.5 Flash pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+6.2points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+5.3points ahead
Every test, side by side
All 24 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingQwen3.5 27B
- Terminal-Bench HardQwen3.5 27B by 1913.632.6+19
- SWE-bench VerifiedQwen3.5 27B by 1260.472.4+12
- SciCodeQwen3.5 27B by 10.429.139.5+10.4
ReasoningQwen3.5 27B
- Humanity's Last ExamQwen3.5 27B by 11.812.123.9+11.8
- GPQA DiamondQwen3.5 27B by 6.87985.8+6.8
- CritPttie1.10.9tie
FactsGemini 2.5 Flash
- AA-Omniscience · Non-hallucinationGemini 2.5 Flash by 6.224.718.5+6.2
- AA-Omniscience · AccuracyGemini 2.5 Flash by 5.325.920.7+5.3
- Vectara HHEM hallucination ratelower is betterGemini 2.5 Flash by 4.37.812.1+4.3
Images and chartsQwen3.5 27B
- MMMU-ProQwen3.5 27B by 5.969.175+5.9
- MMMUQwen3.5 27B by 2.679.782.3+2.6
- LMArena · Visiontie12351241tie
Following instructionsQwen3.5 27B
- IFBenchQwen3.5 27B by 25.350.375.6+25.3
Other results9 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 27B by 62.331.693.9+62.3
- AA Agentic IndexQwen3.5 27B by 35.818.854.6+35.8
- HMMT Feb. 2025Qwen3.5 27B by 27.864.292+27.8
- AA-OmniscienceGemini 2.5 Flash by 14.2-29.8-44+14.2
- Artificial Analysis Coding IndexQwen3.5 27B by 12.722.234.9+12.7
- AA IntelligenceQwen3.5 27B by 9.813.122.9+9.8
- vectara_factual_consistencyGemini 2.5 Flash by 4.392.287.9+4.3
- vectara_answer_ratetie9999.8tie
- vectara_avg_summary_lengthtie101.594.4tie
Questions people ask
Which is better, Gemini 2.5 Flash or Qwen3.5 27B?
Qwen3.5 27B wins six of the seven areas where both have results: coding, agents, reasoning, images and charts, long documents and following instructions. Gemini 2.5 Flash wins facts.
Which is better for coding?
Qwen3.5 27B. It wins 3 of the 3 coding tests both models report; Gemini 2.5 Flash wins none.
Which is cheaper?
Gemini 2.5 Flash costs $0.30 per million input tokens and $2.50 per million output tokens; Qwen3.5 27B costs $0.30 and $2.40. That makes Qwen3.5 27B about 4% cheaper for the same work.
How do you compare the two?
We use the 24 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 9 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.