GPT-4o vs Nova Pro
Wins 4 of 7 areas
Coding · Facts · Images and charts · Long documents
Wins 1 of 7 areas
Following instructions
GPT-4o wins more areas, narrowly.Nova Pro is cheaper and better at following instructions.
Scores updated · 33 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.
- CodingWriting and fixing software20GPT-4o2 of 2 tests
- Images and chartsUnderstanding pictures, charts and video20GPT-4o2 of 2 tests
- FactsGetting facts right instead of making them up21GPT-4o2 of 3 tests
- Long documentsFinding answers in very long texts10GPT-4o1 of 1 test
- Following instructionsDoing exactly what it is asked01Nova Pro1 of 1 test
- AgentsCarrying out multi-step tasks on its own00Even0 each · 1 tie
- ReasoningHard problems that need careful thinking11Even1 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.
Nova Pro costs 80% 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 GPT-4o pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+39.8points ahead
- Reasons across sets of long documentsAA-LCR+35points ahead
- Code for real scientific research problemsSciCode+12.5points ahead
Where Nova Pro pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+2.1points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+1.4points ahead
Every test, side by side
All 33 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-4o
- SciCodeGPT-4o by 12.533.320.8+12.5
- Terminal-Bench HardGPT-4o by 2.28.36.1+2.2
ReasoningEven
- GPQA DiamondGPT-4o by 2.752.649.9+2.7
- Humanity's Last ExamNova Pro by 1.41.83.2+1.4
- CritPttie00tie
FactsGPT-4o
- AA-Omniscience · Non-hallucinationGPT-4o by 39.862.122.3+39.8
- AA-Omniscience · AccuracyGPT-4o by 6.823.716.9+6.8
- Vectara HHEM hallucination ratelower is betterNova Pro by 4.59.65.1+4.5
Images and chartsGPT-4o
Other results20 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.
- AA-OmniscienceGPT-4o by 37.2-10.5-47.7+37.2
- vectara_avg_summary_lengthGPT-4o by 20.486.666.2+20.4
- τ²-Bench Telecom (AA run)GPT-4o by 14.928.914+14.9
- Artificial Analysis Coding IndexGPT-4o by 13.224.211+13.2
- IFEvalNova Pro by 11.18192.1+11.1
- naturalquestions_closedbookGPT-4o by 9.149.640.5+9.1
- vectara_answer_rateNova Pro by 5.593.899.3+5.5
- vectara_factual_consistencyNova Pro by 4.590.494.9+4.5
- GSM8KNova Pro by 3.990.994.8+3.9
- AA Agentic IndexGPT-4o by 3.78.44.7+3.7
- ChartQANova Pro by 3.585.789.2+3.5
- MATHGPT-4o by 3.285.382.1+3.2
- DROPNova Pro by 283.485.4+2
- MMLUNova Pro by 273.875.8+2
- HumanEvalGPT-4o by 1.290.289+1.2
- OpenBookQAtie96.896tie
- DocVQAtie92.893.5tie
- NarrativeQAtie79.579.1tie
- AA Intelligencetie7.37tie
- EgoSchematie72.272.1tie
Questions people ask
Which is better, GPT-4o or Nova Pro?
GPT-4o wins four of the seven areas where both have results: coding, facts, images and charts and long documents. Nova Pro wins following instructions, and costs 80% less. They are level on agents and reasoning.
Which is better for coding?
GPT-4o. It wins 2 of the 2 coding tests both models report; Nova Pro wins none.
Which is cheaper?
GPT-4o costs $5.00 per million input tokens and $15.00 per million output tokens; Nova Pro costs $0.80 and $3.20. That makes Nova Pro about 80% cheaper for the same work.
How do you compare the two?
We use the 33 benchmark tests both models have published scores on. The verdict counts the 13 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 20 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.