Compare · ModelsLive · 2 picked · head to head

Gemini 3 Flash Preview vs Qwen3 235B A22B Thinking 2507

Side by side · benchmarks, pricing, and signals you can act on.

Winner summary

Gemini 3 Flash Preview wins 8 of 8 shared benchmarks. Leads in arena · knowledge · math.

Category leads
arena·Gemini 3 Flash Previewknowledge·Gemini 3 Flash Previewmath·Gemini 3 Flash Previewcoding·Gemini 3 Flash Preview
Hype vs Reality
Gemini 3 Flash Preview
#98 by perf·no signal
QUIET
Qwen3 235B A22B Thinking 2507
#66 by perf·no signal
QUIET
Best value
2.4x better value than Gemini 3 Flash Preview
Gemini 3 Flash Preview
28.1 pts/$
$1.75/M
Qwen3 235B A22B Thinking 2507
68.0 pts/$
$0.82/M
Vendor risk
Google DeepMind logo
Google DeepMind
$4.00T·Tier 1
Low risk
Alibaba Qwen logo
Alibaba (Qwen)
$293.0B·Tier 1
Low risk
Head to head
Gemini 3 Flash PreviewQwen3 235B A22B Thinking 2507
Chatbot Arena Elo · Overall
Gemini 3 Flash Preview leads by +74.1
Gemini 3 Flash Preview
1473.9
Qwen3 235B A22B Thinking 2507
1399.8
Chess Puzzles
Gemini 3 Flash Preview leads by +26.0
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Gemini 3 Flash Preview
38.0
Qwen3 235B A22B Thinking 2507
12.0
FrontierMath-2025-02-28-Private
Gemini 3 Flash Preview leads by +27.2
FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning.
Gemini 3 Flash Preview
35.6
Qwen3 235B A22B Thinking 2507
8.5
FrontierMath-Tier-4-2025-07-01-Private
Gemini 3 Flash Preview leads by +4.1
FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning.
Gemini 3 Flash Preview
4.2
Qwen3 235B A22B Thinking 2507
0.1
GPQA diamond
Gemini 3 Flash Preview leads by +4.2
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Gemini 3 Flash Preview
77.6
Qwen3 235B A22B Thinking 2507
73.4
OTIS Mock AIME 2024-2025
Gemini 3 Flash Preview leads by +6.1
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Gemini 3 Flash Preview
92.8
Qwen3 235B A22B Thinking 2507
86.7
SimpleQA Verified
Gemini 3 Flash Preview leads by +17.3
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Gemini 3 Flash Preview
67.4
Qwen3 235B A22B Thinking 2507
50.1
WeirdML
Gemini 3 Flash Preview leads by +20.6
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Gemini 3 Flash Preview
61.6
Qwen3 235B A22B Thinking 2507
41.0
Full benchmark table
BenchmarkGemini 3 Flash PreviewQwen3 235B A22B Thinking 2507
Chatbot Arena Elo · Overall
1473.91399.8
Chess Puzzles
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
38.012.0
FrontierMath-2025-02-28-Private
FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning.
35.68.5
FrontierMath-Tier-4-2025-07-01-Private
FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning.
4.20.1
GPQA diamond
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
77.673.4
OTIS Mock AIME 2024-2025
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
92.886.7
SimpleQA Verified
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
67.450.1
WeirdML
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
61.641.0
Pricing · per 1M tokens · projected $/mo at 10M tokens
ModelInputOutputContextProjected $/mo
Google DeepMind logoGemini 3 Flash Preview$0.50$3.001.0M tokens (~524 books)$11.25
Alibaba Qwen logoQwen3 235B A22B Thinking 2507$0.15$1.50131K tokens (~66 books)$4.86