Llama 3.1 70B Instruct vs Qwen2.5 Instruct 72B
Wins 1 of 5 areas
Facts
Wins 4 of 5 areas
Coding · Reasoning · Long documents · Following instructions
Qwen2.5 Instruct 72B is the stronger all-rounder.Llama 3.1 70B Instruct is better at facts.
Scores updated · 36 tests both models report · How we compare
Where each one wins
Tests won in each of the five areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- ReasoningHard problems that need careful thinking01Qwen2.5 Instruct 72B1 of 3 tests · 2 ties
- CodingWriting and fixing software01Qwen2.5 Instruct 72B1 of 2 tests · 1 tie
- Long documentsFinding answers in very long texts01Qwen2.5 Instruct 72B1 of 1 test
- Following instructionsDoing exactly what it is asked01Qwen2.5 Instruct 72B1 of 1 test
- FactsGetting facts right instead of making them up20Llama 3.1 70B Instruct2 of 2 tests
Long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
Qwen2.5 Instruct 72B costs 13% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Qwen2.5 Instruct 72B pulls ahead
- Reasons across sets of long documentsAA-LCR+12points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+8.2points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+2.5points ahead
Where Llama 3.1 70B Instruct pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+7.3points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+2.2points ahead
Every test, side by side
All 36 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingQwen2.5 Instruct 72B
- Terminal-Bench HardQwen2.5 Instruct 72B by 1.534.5+1.5
- SciCodetie26.726.7tie
ReasoningQwen2.5 Instruct 72B
- GPQA DiamondQwen2.5 Instruct 72B by 8.240.949.1+8.2
- Humanity's Last Examtie4.53.6tie
- CritPttie00tie
FactsLlama 3.1 70B Instruct
- AA-Omniscience · Non-hallucinationLlama 3.1 70B Instruct by 7.321.814.5+7.3
- AA-Omniscience · AccuracyLlama 3.1 70B Instruct by 2.219.717.5+2.2
Following instructionsQwen2.5 Instruct 72B
- IFBenchQwen2.5 Instruct 72B by 2.534.436.9+2.5
Other results27 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- AlignBenchQwen2.5 Instruct 72B by 75.75.981.6+75.7
- HarmBenchQwen2.5 Instruct 72B by 25.946.972.8+25.9
- Arena HardQwen2.5 Instruct 72B by 25.555.781.2+25.5
- MATHQwen2.5 Instruct 72B by 20.46888.4+20.4
- MBPPQwen2.5 Instruct 72B by 19.668.688.2+19.6
- τ²-Bench Telecom (AA run)Qwen2.5 Instruct 72B by 19.315.234.5+19.3
- AIR-Bench 2024Qwen2.5 Instruct 72B by 16.542.559+16.5
- MGSMLlama 3.1 70B Instruct by 10.786.976.2+10.7
- AA-OmniscienceLlama 3.1 70B Instruct by 10-43.1-53.1+10
- naturalquestions_closedbookLlama 3.1 70B Instruct by 9.345.235.9+9.3
- MMLULlama 3.1 70B Instruct by 98677+9
- simple_safety_testsQwen2.5 Instruct 72B by 7.592.5100+7.5
- anthropic_red_teamQwen2.5 Instruct 72B by 6.493.299.6+6.4
- HumanEvalQwen2.5 Instruct 72B by 6.180.586.6+6.1
- MMLU-ProQwen2.5 Instruct 72B by 4.766.471.1+4.7
- IFEvalLlama 3.1 70B Instruct by 3.487.584.1+3.4
- XSTestQwen2.5 Instruct 72B by 3.494.597.9+3.4
- WinoGrandeLlama 3.1 70B Instruct by 385.382.3+3
- DROPLlama 3.1 70B Instruct by 2.979.676.7+2.9
- NarrativeQALlama 3.1 70B Instruct by 2.777.274.5+2.7
- OpenBookQAQwen2.5 Instruct 72B by 2.493.896.2+2.4
- GSM8KQwen2.5 Instruct 72B by 293.895.8+2
- BBHLlama 3.1 70B Instruct by 1.881.679.8+1.8
- MT-BenchQwen2.5 Instruct 72B by 1.28.29.3+1.2
- AA IntelligenceQwen2.5 Instruct 72B by 1.16.67.7+1.1
- Artificial Analysis Coding IndexQwen2.5 Instruct 72B by 110.911.9+1
- ARC-Challengetie94.894.5tie
Questions people ask
Which is better, Llama 3.1 70B Instruct or Qwen2.5 Instruct 72B?
Qwen2.5 Instruct 72B wins four of the five areas where both have results: coding, reasoning, long documents and following instructions. Llama 3.1 70B Instruct wins facts.
Which is better for coding?
Qwen2.5 Instruct 72B. It wins 1 of the 2 coding tests both models report; Llama 3.1 70B Instruct wins none, and 1 is a tie.
Which is cheaper?
Llama 3.1 70B Instruct costs $0.56 per million input tokens and $0.56 per million output tokens; Qwen2.5 Instruct 72B costs $0.47 and $0.49. That makes Qwen2.5 Instruct 72B about 13% cheaper for the same work.
How do you compare the two?
We use the 36 benchmark tests both models have published scores on. The verdict counts the 9 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 27 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.