Skip to main content

Llama 3.1 70B Instruct vs Qwen2.5 72B Instruct

SpecLlama 3.1 70B InstructQwen2.5 72B Instruct
Respan Index10.6 (#171)10.0 (#172)
Blended price per 1M$0.4$0.37
Context window128K131K
Input / output per 1M$0.4 / $0.4$0.36 / $0.4
Max output-8K
Inputtexttext
Released2024-07-232024-09-19

Respan Index by category

Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.

CategoryLlama 3.1 70B InstructQwen2.5 72B Instruct
Coding0.3 Llama 3.1 70B Instruct0.9 Qwen2.5 72B Instruct
Agents6.6 Llama 3.1 70B Instruct2.0 Qwen2.5 72B Instruct
Reasoning15.7 Llama 3.1 70B Instruct16.6 Qwen2.5 72B Instruct
Knowledge & language3.4 Llama 3.1 70B Instruct3.0* Qwen2.5 72B Instruct
Human preference31.0 Llama 3.1 70B Instruct32.3 Qwen2.5 72B Instruct

* Estimated: no published scores in that category. How the Respan Index works

Benchmarks both report

Qwen2.5 72B Instruct leads on 5 of the 6 benchmarks both report. Higher is better. Each score links to where it was published.

BenchmarkLlama 3.1 70B InstructQwen2.5 72B Instruct
LMArena Elo12931303
GPQA Diamond44.2%49.1%
MATH Level 5 (Epoch AI)36.7%63.2%
OTIS Mock AIME 2024-2025 (Epoch AI)3.6%8.1%
WeirdML (Håvard Tveit Ihle)9%16%
BALROG (BALROG)27.9%16.2%

FAQ

Is Llama 3.1 70B Instruct better than Qwen2.5 72B Instruct?

Llama 3.1 70B Instruct scores 10.6 on the Respan Index (#171), ahead of Qwen2.5 72B Instruct at 10.0 (#172). Qwen2.5 72B Instruct leads on 5 of the 6 benchmarks both report.

Which is cheaper, Llama 3.1 70B Instruct or Qwen2.5 72B Instruct?

Llama 3.1 70B Instruct lists at $0.4 input and $0.4 output per 1M tokens; Qwen2.5 72B Instruct at $0.36 and $0.4. Qwen2.5 72B Instruct is 1.1x cheaper per token at list price.

Which has the larger context window?

Llama 3.1 70B Instruct accepts 128K tokens and Qwen2.5 72B Instruct 131K tokens.

Compare other models

vs

Built for AI agents.
Break less.
Ship more.