GPT-4 vs Qwen2.5 Instruct 72B
Wins 1 of 3 areas
Facts
Wins 2 of 3 areas
Reasoning · Following instructions
Qwen2.5 Instruct 72B is the stronger all-rounder.GPT-4 is better at facts.
Scores updated · 19 tests both models report · How we compare
Where each one wins
Tests won in each of the three 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.
Reasoning 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 99% 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
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+14.2points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+3.7points ahead
Where GPT-4 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+16.1points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+3.5points ahead
Every test, side by side
All 19 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningQwen2.5 Instruct 72B
- GPQA DiamondQwen2.5 Instruct 72B by 14.234.949.1+14.2
FactsGPT-4
- AA-Omniscience · Non-hallucinationGPT-4 by 16.130.614.5+16.1
- AA-Omniscience · AccuracyGPT-4 by 3.52117.5+3.5
Following instructionsQwen2.5 Instruct 72B
- IFBenchQwen2.5 Instruct 72B by 3.733.236.9+3.7
Other results15 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.
- MATHQwen2.5 Instruct 72B by 45.942.588.4+45.9
- HumanEvalQwen2.5 Instruct 72B by 19.66786.6+19.6
- AA-OmniscienceGPT-4 by 19.3-33.8-53.1+19.3
- HellaSwagGPT-4 by 10.595.384.8+10.5
- naturalquestions_closedbookGPT-4 by 9.845.735.9+9.8
- AIR-Bench 2024GPT-4 by 5.264.259+5.2
- WinoGrandeGPT-4 by 5.287.582.3+5.2
- DROPGPT-4 by 4.280.976.7+4.2
- MMLUQwen2.5 Instruct 72B by 3.573.577+3.5
- GSM8KQwen2.5 Instruct 72B by 2.693.295.8+2.6
- NarrativeQAGPT-4 by 2.376.874.5+2.3
- MGSMQwen2.5 Instruct 72B by 1.774.576.2+1.7
- Artificial Analysis Coding IndexGPT-4 by 1.213.111.9+1.2
- AA IntelligenceQwen2.5 Instruct 72B by 16.77.7+1
- OpenBookQAtie9696.2tie
Questions people ask
Which is better, GPT-4 or Qwen2.5 Instruct 72B?
Qwen2.5 Instruct 72B wins two of the three areas where both have results: reasoning and following instructions. GPT-4 wins facts.
Which is cheaper?
GPT-4 costs $30.00 per million input tokens and $60.00 per million output tokens; Qwen2.5 Instruct 72B costs $0.47 and $0.49. That makes Qwen2.5 Instruct 72B about 99% cheaper for the same work.
How do you compare the two?
We use the 19 benchmark tests both models have published scores on. The verdict counts the 4 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 15 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.