Compare · ModelsLive · 2 picked · head to head
Claude 3.5 Sonnet vs Qwen3 14B
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Qwen3 14B wins on 2/3 benchmarks
Qwen3 14B wins 2 of 3 shared benchmarks. Leads in knowledge · math.
Category leads
general·Claude 3.5 Sonnetknowledge·Qwen3 14Bmath·Qwen3 14B
Hype vs Reality
Attention vs performance
Claude 3.5 Sonnet
#190 by perf·no signal
Qwen3 14B
#187 by perf·#2 by attention
Vendor risk
Who is behind the model
Anthropic
$965.0B·Tier 1
Alibaba (Qwen)
$293.0B·Tier 1
Head to head
3 benchmarks · 2 models
Claude 3.5 SonnetQwen3 14B
Dtbench
Claude 3.5 Sonnet leads by +6.3
Claude 3.5 Sonnet
46.3
Qwen3 14B
40.0
GPQA diamond
Qwen3 14B leads by +13.0
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Claude 3.5 Sonnet
38.7
Qwen3 14B
51.7
OTIS Mock AIME 2024-2025
Qwen3 14B leads by +59.9
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Claude 3.5 Sonnet
6.4
Qwen3 14B
66.4
Full benchmark table
| Benchmark | Claude 3.5 Sonnet | Qwen3 14B |
|---|---|---|
Dtbench | 46.3 | 40.0 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 38.7 | 51.7 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 6.4 | 66.4 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| — | — | — | — | |
| $0.12 | $0.24 | 131K tokens (~66 books) | $1.50 |