Compare · ModelsLive · 3 picked · head to head

Llama 3.1 70B Instruct vs Llama 3.3 70B Instruct vs Qwen2.5 72B Instruct

Side by side · benchmarks, pricing, and signals you can act on.

CiteAdd another

Winner summary

Qwen2.5 72B Instruct wins on 11/18 benchmarks

Qwen2.5 72B Instruct wins 11 of 18 shared benchmarks. Leads in coding · knowledge · general.

Llama 3.1 70B Instruct

3 / 18

Llama 3.3 70B Instruct

4 / 18

Qwen2.5 72B Instruct

11 / 18

Category leads

coding·Qwen2.5 72B Instructarena·Llama 3.3 70B Instructknowledge·Qwen2.5 72B Instructgeneral·Qwen2.5 72B Instructlanguage·Llama 3.3 70B Instructmath·Qwen2.5 72B Instructreasoning·Llama 3.1 70B Instructagentic·Llama 3.1 70B Instruct

Hype vs Reality

Attention vs performance

Llama 3.1 70B Instruct

#216 by perf·#18 by attention

QUIET

Llama 3.3 70B Instruct

#218 by perf·#18 by attention

QUIET

Qwen2.5 72B Instruct

#128 by perf·#2 by attention

DESERVED

See full mindshare →

Best value

Llama 3.3 70B Instruct

1.3x better value than Qwen2.5 72B Instruct

Llama 3.1 70B Instruct

90.7 pts/$

$0.40/M

Llama 3.3 70B Instruct

171.9 pts/$

$0.21/M

Qwen2.5 72B Instruct

129.5 pts/$

$0.38/M

Explore pricing →

Vendor risk

Who is behind the model

Meta AI

$1.87T·Tier 1

Low risk

Meta AI

$1.87T·Tier 1

Low risk

Alibaba (Qwen)

$293.0B·Tier 1

Low risk

See the AI economy →

Head to head

18 benchmarks · 3 models

Llama 3.1 70B InstructLlama 3.3 70B InstructQwen2.5 72B Instruct

Aider · Code Editing

Qwen2.5 72B Instruct leads by +6.0

Llama 3.1 70B Instruct

58.6

Llama 3.3 70B Instruct

59.4

Qwen2.5 72B Instruct

65.4

Chatbot Arena Elo · Overall

Llama 3.3 70B Instruct leads by +15.0

Llama 3.1 70B Instruct

1293.3

Llama 3.3 70B Instruct

1317.8

Qwen2.5 72B Instruct

1302.8

Balrog

Llama 3.1 70B Instruct leads by +4.9

Balrog · benchmarks AI agents on text-based adventure games, testing language understanding, strategic planning, and long-horizon reasoning.

Llama 3.1 70B Instruct

27.9

Llama 3.3 70B Instruct

23.0

Qwen2.5 72B Instruct

16.2

Dtbench

Qwen2.5 72B Instruct leads by +4.9

Llama 3.1 70B Instruct

33.3

Llama 3.3 70B Instruct

32.5

Qwen2.5 72B Instruct

38.2

GPQA diamond

Qwen2.5 72B Instruct leads by +2.3

Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.

Llama 3.1 70B Instruct

25.6

Llama 3.3 70B Instruct

29.9

Qwen2.5 72B Instruct

32.2

BBH (HuggingFace)

Qwen2.5 72B Instruct leads by +5.3

Llama 3.1 70B Instruct

55.9

Llama 3.3 70B Instruct

56.6

Qwen2.5 72B Instruct

61.9

GPQA

Qwen2.5 72B Instruct leads by +2.5

Llama 3.1 70B Instruct

14.2

Llama 3.3 70B Instruct

10.5

Qwen2.5 72B Instruct

16.7

IFEval

Llama 3.3 70B Instruct leads by +3.3

Llama 3.1 70B Instruct

86.7

Llama 3.3 70B Instruct

90.0

Qwen2.5 72B Instruct

86.4

MATH Level 5

Qwen2.5 72B Instruct leads by +11.5

Llama 3.1 70B Instruct

38.1

Llama 3.3 70B Instruct

48.3

Qwen2.5 72B Instruct

59.8

MMLU-PRO

Qwen2.5 72B Instruct leads by +3.3

Llama 3.1 70B Instruct

47.9

Llama 3.3 70B Instruct

48.1

Qwen2.5 72B Instruct

51.4

MUSR

Llama 3.1 70B Instruct leads by +2.1

Llama 3.1 70B Instruct

17.7

Llama 3.3 70B Instruct

15.6

Qwen2.5 72B Instruct

11.7

Lmca

Llama 3.3 70B Instruct leads by +3.1

Llama 3.1 70B Instruct

17.5

Llama 3.3 70B Instruct

20.6

Qwen2.5 72B Instruct

15.8

MATH level 5

Qwen2.5 72B Instruct leads by +21.6

MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics.

Llama 3.1 70B Instruct

36.7

Llama 3.3 70B Instruct

41.6

Qwen2.5 72B Instruct

63.2

MMLU

Llama 3.3 70B Instruct leads by +1.3

Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge.

Llama 3.1 70B Instruct

73.5

Llama 3.3 70B Instruct

81.7

Qwen2.5 72B Instruct

80.4

OTIS Mock AIME 2024-2025

Qwen2.5 72B Instruct leads by +2.9

OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.

Llama 3.1 70B Instruct

3.5

Llama 3.3 70B Instruct

5.0

Qwen2.5 72B Instruct

8.0

WeirdML

Qwen2.5 72B Instruct leads by +1.5

WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.

Llama 3.1 70B Instruct

9.0

Llama 3.3 70B Instruct

14.4

Qwen2.5 72B Instruct

16.0

CMMLU

Qwen2.5 72B Instruct leads by +21.3

Llama 3.1 70B Instruct

64.4

Qwen2.5 72B Instruct

85.7

The Agent Company

Llama 3.1 70B Instruct leads by +1.2

The Agent Company · tests AI agents on realistic corporate tasks like email management, code review, data analysis, and cross-tool workflows.

Llama 3.1 70B Instruct

6.9

Qwen2.5 72B Instruct

5.7

Full benchmark table

Benchmark	Llama 3.1 70B Instruct	Llama 3.3 70B Instruct	Qwen2.5 72B Instruct
Aider · Code Editing	58.6	59.4	65.4
Chatbot Arena Elo · Overall	1293.3	1317.8	1302.8
Balrog Balrog · benchmarks AI agents on text-based adventure games, testing language understanding, strategic planning, and long-horizon reasoning.	27.9	23.0	16.2
Dtbench	33.3	32.5	38.2
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.	25.6	29.9	32.2
BBH (HuggingFace)	55.9	56.6	61.9
GPQA	14.2	10.5	16.7
IFEval	86.7	90.0	86.4
MATH Level 5	38.1	48.3	59.8
MMLU-PRO	47.9	48.1	51.4
MUSR	17.7	15.6	11.7
Lmca	17.5	20.6	15.8
MATH level 5 MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics.	36.7	41.6	63.2
MMLU Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge.	73.5	81.7	80.4
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.	3.5	5.0	8.0
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.	9.0	14.4	16.0
CMMLU	64.4	—	85.7
The Agent Company The Agent Company · tests AI agents on realistic corporate tasks like email management, code review, data analysis, and cross-tool workflows.	6.9	—	5.7

Pricing · per 1M tokens · projected $/mo at 10M tokens

Model	Input	Output	Context	Projected $/mo
Llama 3.1 70B Instruct	$0.40	$0.40	131K tokens (~66 books)	$4.00
Llama 3.3 70B Instruct	$0.10	$0.32	131K tokens (~66 books)	$1.55
Qwen2.5 72B Instruct	$0.36	$0.40	33K tokens (~16 books)	$3.70

People also compared

GPT-5.5 Pro vs Llama 3.1 70B Instruct GPT-5.5 vs Llama 3.1 70B Instruct Claude Opus 5.5 vs Llama 3.1 70B Instruct Claude Mythos Preview vs Llama 3.1 70B Instruct DeepSeek V3.2 Speciale vs Llama 3.1 70B Instruct DeepSeek-V2 (MoE-236B, May 2024) vs Llama 3.1 70B Instruct Grok 3 Beta vs Llama 3.1 70B Instruct Llama 3.1 70B Instruct vs MiniMax M2