DeepSeek-V4.1-Flash vs DeepSeek-V4-Pro
Wins 6 of 7 areas
Coding · Agents · Reasoning · Math · Long documents · Following instructions
Wins 1 of 7 areas
Facts
DeepSeek-V4.1-Flash is the stronger all-rounder.DeepSeek-V4-Pro is better at facts.
Scores updated · 36 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 software50DeepSeek-V4.1-Flash5 of 5 tests
- ReasoningHard problems that need careful thinking50DeepSeek-V4.1-Flash5 of 5 tests
- AgentsCarrying out multi-step tasks on its own30DeepSeek-V4.1-Flash3 of 3 tests
- MathCompetition and research-level math30DeepSeek-V4.1-Flash3 of 3 tests
- Long documentsFinding answers in very long texts10DeepSeek-V4.1-Flash1 of 1 test
- Following instructionsDoing exactly what it is asked10DeepSeek-V4.1-Flash1 of 1 test
- FactsGetting facts right instead of making them up12DeepSeek-V4-Pro2 of 3 tests
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-V4.1-Flash costs 72% 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 DeepSeek-V4.1-Flash pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+43.9points ahead
- Fixes real GitHub issues, working as an agentLiveBench · Agentic Coding+34.7points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+26.6points ahead
Where DeepSeek-V4-Pro pulls ahead
- Short factual questions, answered correctlySimpleQA Verified+3.3points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+2.4points ahead
Every test, side by side
All 36 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingDeepSeek-V4.1-Flash
- LiveBench · Agentic CodingDeepSeek-V4.1-Flash by 34.777.342.6+34.7
- Terminal-Bench 2.1DeepSeek-V4.1-Flash by 26.690.664+26.6
- LMArena · WebDevDeepSeek-V4.1-Flash by 155 rating points16191464+155 rating
- LiveBench · CodingDeepSeek-V4.1-Flash by 108070+10
- SciCodeDeepSeek-V4.1-Flash by 1.151.950.8+1.1
AgentsDeepSeek-V4.1-Flash
- GDPValDeepSeek-V4.1-Flash by 22.25532.8+22.2
- Terminal-Bench 4.0DeepSeek-V4.1-Flash by 12.226.814.6+12.2
- AA IT-Bench SREDeepSeek-V4.1-Flash by 8.646.938.3+8.6
ReasoningDeepSeek-V4.1-Flash
- SimpleBenchDeepSeek-V4.1-Flash by 15.866.750.9+15.8
- LiveBench · ReasoningDeepSeek-V4.1-Flash by 486.782.7+4
- GPQA DiamondDeepSeek-V4.1-Flash by 2.190.988.8+2.1
- Humanity's Last ExamDeepSeek-V4.1-Flash by 1.739.237.5+1.7
- CritPtDeepSeek-V4.1-Flash by 1.414.312.9+1.4
FactsDeepSeek-V4-Pro
- AA-Omniscience · AccuracyDeepSeek-V4.1-Flash by 3.446.443+3.4
- SimpleQA VerifiedDeepSeek-V4-Pro by 3.342.946.2+3.3
- AA-Omniscience · Non-hallucinationDeepSeek-V4-Pro by 2.43.55.9+2.4
MathDeepSeek-V4.1-Flash
- FrontierMath Tier 4DeepSeek-V4.1-Flash by 43.946.32.4+43.9
- FrontierMath Tiers 1-3 (v2)DeepSeek-V4.1-Flash by 25.971.245.3+25.9
- LiveBench · MathematicsDeepSeek-V4.1-Flash by 2.693.390.7+2.6
Long documentsDeepSeek-V4.1-Flash
- AA-LCRDeepSeek-V4.1-Flash by 9.38474.7+9.3
Following instructionsDeepSeek-V4.1-Flash
- LiveBench · Instruction FollowingDeepSeek-V4.1-Flash by 7.67062.4+7.6
Other results15 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.
- Program BenchDeepSeek-V4-Pro by 27.520.347.8+27.5
- NL2RepoDeepSeek-V4.1-Flash by 25.56438.5+25.5
- MathArena Apex (Pass@1)DeepSeek-V4-Pro by 24.665.690.2+24.6
- Terminal-Bench 4.0DeepSeek-V4.1-Flash by 18.831.212.4+18.8
- Terminal-bench 3.0DeepSeek-V4.1-Flash by 18.23011.8+18.2
- AA IntelligenceDeepSeek-V4.1-Flash by 9.139.530.4+9.1
- Agents' Last ExamDeepSeek-V4.1-Flash by 6.131.825.7+6.1
- AA-OmniscienceDeepSeek-V4.1-Flash by 5.4-5.3-10.7+5.4
- CyberGymDeepSeek-V4.1-Flash by 4.888.183.3+4.8
- livebench_data_analysisDeepSeek-V4.1-Flash by 4.879.374.5+4.8
- vectara_avg_summary_lengthDeepSeek-V4-Pro by 45.6 rating points108.2153.8+45.6 rating
- vectara_factual_consistencyDeepSeek-V4.1-Flash by 4.495.891.4+4.4
- HLE (with tools)DeepSeek-V4.1-Flash by 3.963.960+3.9
- livebench_languageDeepSeek-V4.1-Flash by 3.181.278.1+3.1
- vectara_answer_rateDeepSeek-V4.1-Flash by 2.799.997.2+2.7
Questions people ask
Which is better, DeepSeek-V4.1-Flash or DeepSeek-V4-Pro?
DeepSeek-V4.1-Flash wins six of the seven areas where both have results: coding, agents, reasoning, math, long documents and following instructions. DeepSeek-V4-Pro wins facts.
Which is better for coding?
DeepSeek-V4.1-Flash. It wins 5 of the 5 coding tests both models report; DeepSeek-V4-Pro wins none.
Which is cheaper?
DeepSeek-V4.1-Flash costs $0.30 per million input tokens and $1.20 per million output tokens; DeepSeek-V4-Pro costs $1.32 and $3.96. That makes DeepSeek-V4.1-Flash about 72% cheaper for the same work.
How do you compare the two?
We use the 36 benchmark tests both models have published scores on. The verdict counts the 21 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 15 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.