DeepSeek-V3.2 vs o1
Wins 3 of 6 areas
Coding · Reasoning · Long documents
Wins 3 of 6 areas
Agents · Facts · Following instructions
The two are evenly matched.DeepSeek-V3.2 is better at coding and reasoning; o1 at facts and agents.
Scores updated · 18 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 software30DeepSeek-V3.23 of 3 tests
- ReasoningHard problems that need careful thinking30DeepSeek-V3.23 of 3 tests
- Long documentsFinding answers in very long texts10DeepSeek-V3.21 of 1 test
- FactsGetting facts right instead of making them up03o13 of 3 tests
- AgentsCarrying out multi-step tasks on its own01o11 of 1 test
- Following instructionsDoing exactly what it is asked01o11 of 1 test
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.
DeepSeek-V3.2 costs 99% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where DeepSeek-V3.2 pulls ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+31.8points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+22.7points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+17.6points ahead
Where o1 pulls ahead
- Short factual questions, answered correctlySimpleQA Verified+13.6points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+13.1points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+9.6points ahead
Every test, side by side
All 18 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingDeepSeek-V3.2
- SWE-bench VerifiedDeepSeek-V3.2 by 31.873.141.3+31.8
- Terminal-Bench HardDeepSeek-V3.2 by 22.735.612.9+22.7
- SciCodeDeepSeek-V3.2 by 3.138.935.8+3.1
ReasoningDeepSeek-V3.2
- Humanity's Last ExamDeepSeek-V3.2 by 17.624.67+17.6
- GPQA DiamondDeepSeek-V3.2 by 9.38474.7+9.3
- CritPtDeepSeek-V3.2 by 2.62.90.3+2.6
Factso1
- SimpleQA Verifiedo1 by 13.627.541.1+13.6
- AA-Omniscience · Non-hallucinationo1 by 13.117.330.4+13.1
- AA-Omniscience · Accuracyo1 by 1.53334.5+1.5
Other results6 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)DeepSeek-V3.2 by 2890.662.6+28
- AA Agentic Indexo1 by 12.818.331.1+12.8
- AA-Omniscienceo1 by 11.5-22.5-11.1+11.5
- AIME 2025DeepSeek-V3.2 by 11.493.181.7+11.4
- AA IntelligenceDeepSeek-V3.2 by 6.321.515.2+6.3
- Artificial Analysis Coding IndexDeepSeek-V3.2 by 4.544.239.7+4.5
Questions people ask
Which is better, DeepSeek-V3.2 or o1?
DeepSeek-V3.2 and o1 each win three of the six areas where both have results. DeepSeek-V3.2 wins coding, reasoning and long documents; o1 wins agents, facts and following instructions. DeepSeek-V3.2 costs 99% less.
Which is better for coding?
DeepSeek-V3.2. It wins 3 of the 3 coding tests both models report; o1 wins none.
Which is cheaper?
DeepSeek-V3.2 costs $0.28 per million input tokens and $0.42 per million output tokens; o1 costs $15.00 and $60.00. That makes DeepSeek-V3.2 about 99% cheaper for the same work.
How do you compare the two?
We use the 18 benchmark tests both models have published scores on. The verdict counts the 12 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 6 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.