EXAONE 4.5 33B vs GPT-4o
Wins 4 of 7 areas
Coding · Reasoning · Images and charts · Following instructions
Wins 2 of 7 areas
Facts · Long documents
EXAONE 4.5 33B wins more areas, narrowly.GPT-4o is better at facts.
Scores updated · 26 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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.
- Images and chartsUnderstanding pictures, charts and video40EXAONE 4.5 33B4 of 4 tests
- ReasoningHard problems that need careful thinking20EXAONE 4.5 33B2 of 3 tests · 1 tie
- CodingWriting and fixing software21EXAONE 4.5 33B2 of 3 tests
- Following instructionsDoing exactly what it is asked10EXAONE 4.5 33B1 of 1 test
- FactsGetting facts right instead of making them up02GPT-4o2 of 2 tests
- Long documentsFinding answers in very long texts01GPT-4o1 of 1 test
- AgentsCarrying out multi-step tasks on its own00Even0 each · 1 tie
Agents, 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.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where EXAONE 4.5 33B pulls ahead
- Recent programming contest problemsLiveCodeBench v6+50.5points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+26.8points ahead
- Math problems shown in pictures and chartsMathVista+23.6points ahead
Where GPT-4o pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+42points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+7.5points ahead
- Code for real scientific research problemsSciCode+5.3points ahead
Every test, side by side
All 26 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingEXAONE 4.5 33B
- LiveCodeBench v6EXAONE 4.5 33B by 50.581.430.9+50.5
- Terminal-Bench HardEXAONE 4.5 33B by 12.220.58.3+12.2
- SciCodeGPT-4o by 5.32833.3+5.3
ReasoningEXAONE 4.5 33B
- GPQA DiamondEXAONE 4.5 33B by 26.879.452.6+26.8
- Humanity's Last ExamEXAONE 4.5 33B by 11.112.91.8+11.1
- CritPttie0.30tie
FactsGPT-4o
- AA-Omniscience · Non-hallucinationGPT-4o by 4220.162.1+42
- AA-Omniscience · AccuracyGPT-4o by 7.516.223.7+7.5
Images and chartsEXAONE 4.5 33B
Following instructionsEXAONE 4.5 33B
- IFBenchEXAONE 4.5 33B by 225836+22
Other results11 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.
- AIME 2025EXAONE 4.5 33B by 81.292.911.7+81.2
- τ²-Bench Telecom (AA run)EXAONE 4.5 33B by 49.278.128.9+49.2
- AA-OmniscienceGPT-4o by 40.2-50.7-10.5+40.2
- τ²-Bench (Retail)EXAONE 4.5 33B by 14.577.963.4+14.5
- Tau2 airlineEXAONE 4.5 33B by 1156.545.5+11
- IFEvalEXAONE 4.5 33B by 8.689.681+8.6
- MMLU-ProEXAONE 4.5 33B by 8.683.374.7+8.6
- AA IntelligenceEXAONE 4.5 33B by 5.913.27.3+5.9
- AI2DGPT-4o by 5.28994.2+5.2
- AA Agentic IndexEXAONE 4.5 33B by 1.39.78.4+1.3
- Artificial Analysis Coding Indextie23.624.2tie
Questions people ask
Which is better, EXAONE 4.5 33B or GPT-4o?
EXAONE 4.5 33B wins four of the seven areas where both have results: coding, reasoning, images and charts and following instructions. GPT-4o wins facts and long documents. They are level on agents.
Which is better for coding?
EXAONE 4.5 33B. It wins 2 of the 3 coding tests both models report; GPT-4o wins 1.
How do you compare the two?
We use the 26 benchmark tests both models have published scores on. The verdict counts the 15 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 11 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.