GPT-4o vs Qwen3 4B 2507 Instruct
Wins 4 of 6 areas
Coding · Facts · Long documents · Following instructions
Wins 1 of 6 areas
Reasoning
GPT-4o is the stronger all-rounder.Qwen3 4B 2507 Instruct is cheaper and better at reasoning.
Scores updated · 23 tests both models report · How we compare
Where each one wins
Tests won in each of the six 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.
- FactsGetting facts right instead of making them up20GPT-4o2 of 2 tests
- CodingWriting and fixing software21GPT-4o2 of 3 tests
- Long documentsFinding answers in very long texts10GPT-4o1 of 1 test
- Following instructionsDoing exactly what it is asked10GPT-4o1 of 1 test
- ReasoningHard problems that need careful thinking01Qwen3 4B 2507 Instruct1 of 3 tests · 2 ties
- 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.
Qwen3 4B 2507 Instruct costs 100% 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
- Reasons across sets of long documentsAA-LCR+44.7points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+43.8points ahead
- Code for real scientific research problemsSciCode+15.2points ahead
Where Qwen3 4B 2507 Instruct pulls ahead
- Recent programming contest problemsLiveCodeBench v6+17.8points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+2.7points ahead
Every test, side by side
All 23 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-4o
- LiveCodeBench v6Qwen3 4B 2507 Instruct by 17.830.948.7+17.8
- SciCodeGPT-4o by 15.233.318.1+15.2
- Terminal-Bench HardGPT-4o by 3.88.34.5+3.8
ReasoningQwen3 4B 2507 Instruct
- Humanity's Last ExamQwen3 4B 2507 Instruct by 2.71.84.5+2.7
- GPQA Diamondtie52.651.7tie
- CritPttie00tie
FactsGPT-4o
- AA-Omniscience · Non-hallucinationGPT-4o by 43.862.118.3+43.8
- AA-Omniscience · AccuracyGPT-4o by 12.823.710.9+12.8
Other results12 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 51.4-10.5-61.9+51.4
- GSM8KGPT-4o by 22.490.968.5+22.4
- MMLU-ProGPT-4o by 22.474.752.3+22.4
- MMMLUGPT-4o by 20.781.460.7+20.7
- Artificial Analysis Coding IndexGPT-4o by 15.224.29+15.2
- MATH-500 (EM)Qwen3 4B 2507 Instruct by 1174.685.6+11
- MGSMGPT-4o by 8.790.581.8+8.7
- IFEvalQwen3 4B 2507 Instruct by 4.68185.6+4.6
- τ²-Bench Telecom (AA run)GPT-4o by 2.328.926.6+2.3
- MMLUGPT-4o by 1.573.872.3+1.5
- AA Intelligencetie7.36.7tie
- AA Agentic Indextie8.48.9tie
Questions people ask
Which is better, GPT-4o or Qwen3 4B 2507 Instruct?
GPT-4o wins four of the six areas where both have results: coding, facts, long documents and following instructions. Qwen3 4B 2507 Instruct wins reasoning, and costs 100% less. They are level on agents.
Which is better for coding?
GPT-4o. It wins 2 of the 3 coding tests both models report; Qwen3 4B 2507 Instruct wins 1.
Which is cheaper?
GPT-4o costs $5.00 per million input tokens and $15.00 per million output tokens; Qwen3 4B 2507 Instruct costs $0.01 and $0.03. That makes Qwen3 4B 2507 Instruct about 100% cheaper for the same work.
How do you compare the two?
We use the 23 benchmark tests both models have published scores on. The verdict counts the 11 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 12 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.