← Back to all models
Q

Qwen3 72B

by Alibaba
Flagship Free & Open Source 🏆 Ranked #18 of 85
84.9
Overall Score
out of 100
About

Alibaba's latest flagship open-source model with a unique dual-mode operation: fast standard responses or an extended thinking mode for harder problems. Qwen3 72B achieves frontier-class mathematics and coding scores as a fully open-source model, rivalling paid APIs.

Key Metrics
Context Window
128K
tokens
Avg Response
900
milliseconds
Input Cost
$0.0
per million tokens
Output Cost
$0.0
per million tokens
Arena ELO
1320
Chatbot Arena rating
MT-Bench
9.0
out of 10
Benchmark Scores
MMLU
88.0%
HumanEval
92.5%
MATH
90.0%
GPQA
65.4%
MT-Bench
90.0/10
Capability Profile
Strengths & Limitations
Strengths
✓ Thinking + standard dual mode ✓ Exceptional mathematics ✓ Open source ✓ Strong coding ✓ Multilingual
Limitations
⚠ Requires significant GPU VRAM ⚠ Thinking mode is slower ⚠ Newer ecosystem
Ideal Use Cases
Mathematics and science Code generation Research Multilingual AI On-premise deployment
Model Details
Provider Alibaba
Released 2025-04-29
Type Free & Open Source
Multimodal No
Tier Flagship
Global rank #18 / 85
Pricing (USD)
Input tokens $0.0/M
Output tokens $0.0/M
Per 1,000 tokens ≈ $0.0000 input / $0.0000 output
All Benchmarks
MMLU 88.0%
HumanEval 92.5%
MATH 90.0%
GPQA 65.4%
MT-Bench 9.0/10
Arena ELO 1320
Compare this model View Rankings

You might also consider