Compare · ModelsLive · 2 picked · head to head
Qwen3 30B A3B vs QwQ 32B
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
QwQ 32B wins on 5/6 benchmarks
QwQ 32B wins 5 of 6 shared benchmarks. Leads in arena · knowledge.
Category leads
arena·QwQ 32Bknowledge·QwQ 32Bmath·Qwen3 30B A3B
Hype vs Reality
Attention vs performance
Qwen3 30B A3B
#199 by perf·#2 by attention
QwQ 32B
#247 by perf·no signal
Best value
Qwen3 30B A3B
1.5x better value than QwQ 32B
Qwen3 30B A3B
124.8 pts/$
$0.31/M
QwQ 32B
84.7 pts/$
$0.36/M
Vendor risk
Who is behind the model
Alibaba (Qwen)
$293.0B·Tier 1
Alibaba (Qwen)
$293.0B·Tier 1
Head to head
6 benchmarks · 2 models
Qwen3 30B A3BQwQ 32B
Chatbot Arena Elo · Overall
QwQ 32B leads by +9.3
Qwen3 30B A3B
1326.5
QwQ 32B
1335.8
Chess Puzzles
QwQ 32B leads by +0.0
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Qwen3 30B A3B
0.0
QwQ 32B
0.0
Fiction.LiveBench
QwQ 32B leads by +42.7
Fiction.LiveBench · a continuously updated benchmark using recently published fiction to test reading comprehension and reasoning, preventing data contamination.
Qwen3 30B A3B
40.6
QwQ 32B
83.3
GPQA diamond
QwQ 32B leads by +4.9
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Qwen3 30B A3B
48.9
QwQ 32B
53.8
Lech Mazur Writing
QwQ 32B leads by +4.9
Lech Mazur Writing · evaluates creative writing ability, assessing prose quality, narrative coherence, and stylistic sophistication.
Qwen3 30B A3B
75.3
QwQ 32B
80.2
OTIS Mock AIME 2024-2025
Qwen3 30B A3B leads by +3.6
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Qwen3 30B A3B
62.7
QwQ 32B
59.1
Full benchmark table
| Benchmark | Qwen3 30B A3B | QwQ 32B |
|---|---|---|
Chatbot Arena Elo · Overall | 1326.5 | 1335.8 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 0.0 | 0.0 |
Fiction.LiveBench Fiction.LiveBench · a continuously updated benchmark using recently published fiction to test reading comprehension and reasoning, preventing data contamination. | 40.6 | 83.3 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 48.9 | 53.8 |
Lech Mazur Writing Lech Mazur Writing · evaluates creative writing ability, assessing prose quality, narrative coherence, and stylistic sophistication. | 75.3 | 80.2 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 62.7 | 59.1 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.12 | $0.50 | 131K tokens (~66 books) | $2.15 | |
| $0.15 | $0.58 | 131K tokens (~66 books) | $2.57 |