DeepSeek-V4.1-Flash vs Gemini 4 Argon
Wins 1 of 6 areas
Long documents
Wins 5 of 6 areas
Coding · Agents · Reasoning · Facts · Math
Gemini 4 Argon is the stronger all-rounder.DeepSeek-V4.1-Flash is cheaper and better at long documents.
Scores updated · 25 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.
- ReasoningHard problems that need careful thinking04Gemini 4 Argon4 of 4 tests
- FactsGetting facts right instead of making them up03Gemini 4 Argon3 of 3 tests
- CodingWriting and fixing software02Gemini 4 Argon2 of 2 tests
- AgentsCarrying out multi-step tasks on its own02Gemini 4 Argon2 of 2 tests
- MathCompetition and research-level math02Gemini 4 Argon2 of 2 tests
- Long documentsFinding answers in very long texts10DeepSeek-V4.1-Flash1 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-V4.1-Flash costs 88% 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 Gemini 4 Argon pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+81.4points ahead
- Solves new puzzle games it has never seenARC-AGI-3+46.1points ahead
- Short factual questions, answered correctlySimpleQA Verified+33.1points ahead
Where DeepSeek-V4.1-Flash pulls ahead
- Reasons across sets of long documentsAA-LCR+4.3points ahead
Every test, side by side
All 25 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 4 Argon
- SciCodeGemini 4 Argon by 9.951.961.8+9.9
- LMArena · WebDevGemini 4 Argon by 60 rating points16191679+60 rating
AgentsGemini 4 Argon
- Terminal-Bench 4.0Gemini 4 Argon by 30.326.857.1+30.3
- GDPValGemini 4 Argon by 1.25556.2+1.2
ReasoningGemini 4 Argon
- ARC-AGI-3Gemini 4 Argon by 46.10.246.3+46.1
- Humanity's Last ExamGemini 4 Argon by 17.939.257.1+17.9
- ARC-AGI-2Gemini 4 Argon by 13.278.191.3+13.2
- CritPtGemini 4 Argon by 12.814.327.1+12.8
FactsGemini 4 Argon
- AA-Omniscience · Non-hallucinationGemini 4 Argon by 81.43.584.9+81.4
- SimpleQA VerifiedGemini 4 Argon by 33.142.976+33.1
- AA-Omniscience · AccuracyGemini 4 Argon by 3.546.449.9+3.5
MathGemini 4 Argon
- FrontierMath Tier 4Gemini 4 Argon by 31.846.378+31.8
- FrontierMath Tiers 1-3 (v2)Gemini 4 Argon by 14.871.286+14.8
Long documentsDeepSeek-V4.1-Flash
- AA-LCRDeepSeek-V4.1-Flash by 4.38479.7+4.3
Other results11 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-OmniscienceGemini 4 Argon by 47.7-5.342.4+47.7
- Terminal-Bench-Science 0.1Gemini 4 Argon by 41.915.757.6+41.9
- Terminal-Bench 4.0Gemini 4 Argon by 26.231.257.4+26.2
- OSWorld 2.0Gemini 4 Argon by 14.954.369.2+14.9
- AA IntelligenceGemini 4 Argon by 13.139.552.6+13.1
- vectara_factual_consistencyDeepSeek-V4.1-Flash by 7.995.887.9+7.9
- Agents' Last ExamGemini 4 Argon by 7.731.839.5+7.7
- vectara_avg_summary_lengthGemini 4 Argon by 70 rating points108.2178.2+70 rating
- ARC-AGI-1Gemini 4 Argon by 593.598.5+5
- DeepSWE 1.1Gemini 4 Argon by 3.774.277.9+3.7
- vectara_answer_ratetie99.999.1tie
Questions people ask
Which is better, DeepSeek-V4.1-Flash or Gemini 4 Argon?
Gemini 4 Argon wins five of the six areas where both have results: coding, agents, reasoning, facts and math. DeepSeek-V4.1-Flash wins long documents, and costs 88% less.
Which is better for coding?
Gemini 4 Argon. It wins 2 of the 2 coding tests both models report; DeepSeek-V4.1-Flash wins none.
Which is cheaper?
DeepSeek-V4.1-Flash costs $0.30 per million input tokens and $1.20 per million output tokens; Gemini 4 Argon costs $2.00 and $10.00. That makes DeepSeek-V4.1-Flash about 88% cheaper for the same work.
How do you compare the two?
We use the 25 benchmark tests both models have published scores on. The verdict counts the 14 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 11 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.