DeepSeek-R1 0528 vs GPT-5
Wins 0 of 6 areas
—
Wins 6 of 6 areas
Coding · Agents · Reasoning · Facts · Long documents · Following instructions
GPT-5 is the stronger all-rounder.DeepSeek-R1 0528 is cheaper.
Scores updated · 37 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.
- CodingWriting and fixing software04GPT-54 of 4 tests
- ReasoningHard problems that need careful thinking04GPT-54 of 4 tests
- FactsGetting facts right instead of making them up03GPT-53 of 3 tests
- AgentsCarrying out multi-step tasks on its own02GPT-52 of 2 tests
- Following instructionsDoing exactly what it is asked02GPT-52 of 2 tests
- Long documentsFinding answers in very long texts01GPT-51 of 1 test
Long documents rests on a single test.
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.
DeepSeek-R1 0528 costs 61% 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-5 pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+46points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+33.5points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+30.3points ahead
Where DeepSeek-R1 0528 pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 37 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5
- SWE-bench VerifiedGPT-5 by 30.344.674.9+30.3
- SWE-bench MultilingualGPT-5 by 24.830.555.3+24.8
- Terminal-Bench HardGPT-5 by 16.715.932.6+16.7
- SciCodeGPT-5 by 2.640.342.9+2.6
ReasoningGPT-5
- SimpleBenchGPT-5 by 15.940.856.7+15.9
- Humanity's Last ExamGPT-5 by 12.715.828.5+12.7
- CritPtGPT-5 by 4.31.45.7+4.3
- GPQA DiamondGPT-5 by 4.181.385.4+4.1
FactsGPT-5
- SimpleQA VerifiedGPT-5 by 22.727.450.1+22.7
- AA-Omniscience · AccuracyGPT-5 by 9.830.540.3+9.8
- AA-Omniscience · Non-hallucinationGPT-5 by 1.216.617.8+1.2
Following instructionsGPT-5
- IFBenchGPT-5 by 33.539.673.1+33.5
- Multi-ChallengeGPT-5 by 24.64569.6+24.6
Other results21 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.
- τ²-Bench Telecom (AA run)GPT-5 by 48.336.584.8+48.3
- ARC-AGI-1GPT-5 by 44.521.265.7+44.5
- HarmBenchGPT-5 by 4354.697.6+43
- Terminal-BenchGPT-5 by 38.15.743.8+38.1
- imo_2025GPT-5 by 31.26.838.1+31.2
- BrowseComp-ZHGPT-5 by 27.335.763+27.3
- AA-OmniscienceGPT-5 by 18.7-27.4-8.7+18.7
- Aider-PolyglotGPT-5 by 16.471.688+16.4
- HMMT 2025GPT-5 by 13.979.493.3+13.9
- Artificial Analysis Coding IndexGPT-5 by 13.82437.8+13.8
- HMMT Feb. 2025GPT-5 by 11.676.788.3+11.6
- LiveCodeBenchGPT-5 by 10.273.383.5+10.2
- AA IntelligenceGPT-5 by 9.913.123+9.9
- AIME 2025GPT-5 by 7.187.594.6+7.1
- AA Agentic IndexGPT-5 by 5.720.826.5+5.7
- MMLU-ProGPT-5 by 2.18587.1+2.1
- MMLU-ReduxGPT-5 by 1.993.495.3+1.9
- XSTestDeepSeek-R1 0528 by 1.798.897.1+1.7
- simple_safety_testsGPT-5 by 1.598.399.8+1.5
- bbqtie96.596.8tie
- anthropic_red_teamtie99.199.1tie
Questions people ask
Which is better, DeepSeek-R1 0528 or GPT-5?
GPT-5 wins all six areas where both have results: coding, agents, reasoning, facts, long documents and following instructions. DeepSeek-R1 0528 wins none, but costs 61% less.
Which is better for coding?
GPT-5. It wins 4 of the 4 coding tests both models report; DeepSeek-R1 0528 wins none.
Which is cheaper?
DeepSeek-R1 0528 costs $1.35 per million input tokens and $3.00 per million output tokens; GPT-5 costs $1.25 and $10.00. That makes DeepSeek-R1 0528 about 61% cheaper for the same work.
How do you compare the two?
We use the 37 benchmark tests both models have published scores on. The verdict counts the 16 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 21 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.