Key Specifications

SpecificationQwen1.5 72BQwen2 72B
Vendoralibabaalibaba
Version1.5-72b2-72b
Release Date2024-02-252024-06-07
Context Window32768 tokens131072 tokens
Input Modalitiestexttext
Output Modalitiestexttext
LicenseQwen LicenseQwen License
SOC2
HIPAA
GDPR
ISO 27001

Benchmark Results

BenchmarkQwen1.5 72BQwen2 72BWinner
ARC92.694.4Qwen2 72B
BBH76.583.4Qwen2 72B
GPQA39.853.6Qwen2 72B
GSM8K80.987Qwen2 72B
HUMANEVAL65.276.5Qwen2 72B
IFEVAL76.476.9Qwen2 72B
MATH37.671.9Qwen2 72B
MMLU81.380.2Qwen1.5 72B
MUSR50.868.9Qwen2 72B
WINOGRANDE83.182.5Qwen1.5 72B

Pricing Comparison

Tier (per Mtok)Qwen1.5 72BQwen2 72B
Input$0.5$0.5
Output$0.8$0.8
Cache Read$0$0
Cache Write$0$0

Qwen1.5 72B vs Qwen2 72B

Model Overview

Qwen1.5 72B and Qwen2 72B are both notable options in the AI model market. This page compares their benchmarks, pricing, and compliance.

Key Specifications

VendorRelease DateContext WindowLicense
Alibaba / Alibaba2024-02-25 / 2024-06-0732K / 131KQwen License / Qwen License

Benchmark Performance

BenchmarkQwen1.5 72BQwen2 72BWinner
ARC92.694.4B
BBH (BIG-Bench Hard)76.583.4B
GPQA39.853.6B
GSM8K (Grade School Math 8K)80.987.0B
HumanEval65.276.5B
IFEval76.476.9B
MATH37.671.9B
MMLU (Massive Multitask Language Understanding)81.380.2A
MUSR50.868.9B
WinoGrande83.182.5A

Pricing Comparison

InputOutputCache ReadCache Write
— / —— / —— / —— / —

per million tokens — A / B

Strengths & Weaknesses

Qwen1.5 72B

  • ✅ MMLU score 81.3, strong knowledge reasoning.
  • ⚠️ Proprietary, not self-hostable.

Qwen2 72B

  • ✅ MMLU score 80.2, strong knowledge reasoning.
  • ✅ GSM8K 87.0, robust math reasoning.
  • ⚠️ Proprietary, not self-hostable.

Editor’s Take

Qwen1.5 72B and Qwen2 72B each have their strengths. Choose based on workload (code, long context, vision), referencing the tables above.

FAQ

Which model is better for coding tasks?

Refer to the HumanEval benchmark table; the model with a higher score is better suited for coding tasks.

Which model is cheaper?

Refer to the pricing comparison table above; the model with lower input/output prices is more cost-effective.

Which has a longer context window?

Refer to the key specifications table; the model with a larger context window is better for long documents.

References

Editor's Take

See Editor's Take section.