DeepSeek-V3.2-Speciale vs Gemini 3 Flash Preview
Wins 0 of 6 areas
—
Wins 5 of 6 areas
Coding · Agents · Reasoning · Long documents · Following instructions
Gemini 3 Flash Preview is the stronger all-rounder.DeepSeek-V3.2-Speciale is cheaper.
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.
- ReasoningHard problems that need careful thinking04Gemini 3 Flash Preview4 of 4 tests
- CodingWriting and fixing software03Gemini 3 Flash Preview3 of 3 tests
- AgentsCarrying out multi-step tasks on its own01Gemini 3 Flash Preview1 of 1 test
- Long documentsFinding answers in very long texts01Gemini 3 Flash Preview1 of 1 test
- Following instructionsDoing exactly what it is asked01Gemini 3 Flash Preview1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
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-Speciale costs 79% 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 3 Flash Preview pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+15.6points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+14.1points ahead
- Common-sense trick questionsSimpleBench+8.5points ahead
Where DeepSeek-V3.2-Speciale pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+3.8points 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.
CodingGemini 3 Flash Preview
- SciCodeGemini 3 Flash Preview by 6.64450.6+6.6
- SWE-bench VerifiedGemini 3 Flash Preview by 4.973.178+4.9
- Terminal-Bench HardGemini 3 Flash Preview by 3.834.838.6+3.8
ReasoningGemini 3 Flash Preview
- SimpleBenchGemini 3 Flash Preview by 8.552.661.1+8.5
- Humanity's Last ExamGemini 3 Flash Preview by 7.928.736.6+7.9
- GPQA DiamondGemini 3 Flash Preview by 2.787.189.8+2.7
- CritPtGemini 3 Flash Preview by 1.27.48.6+1.2
FactsEven
- AA-Omniscience · AccuracyGemini 3 Flash Preview by 15.637.853.4+15.6
- AA-Omniscience · Non-hallucinationDeepSeek-V3.2-Speciale by 3.810.87+3.8
Long documentsGemini 3 Flash Preview
- AA-LCRGemini 3 Flash Preview by 87078+8
Following instructionsGemini 3 Flash Preview
- IFBenchGemini 3 Flash Preview by 14.163.978+14.1
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.
- τ²-Bench Telecom (AA run)Gemini 3 Flash Preview by 80.4080.4+80.4
- AA Agentic IndexGemini 3 Flash Preview by 49.7049.7+49.7
- AA-OmniscienceGemini 3 Flash Preview by 27.8-17.710.1+27.8
- ToolathlonGemini 3 Flash Preview by 14.235.249.4+14.2
- AA IntelligenceGemini 3 Flash Preview by 11.814.526.3+11.8
- t2-benchGemini 3 Flash Preview by 9.980.390.2+9.9
- Artificial Analysis Coding IndexGemini 3 Flash Preview by 4.737.942.6+4.7
- AIME 2025Gemini 3 Flash Preview by 3.79699.7+3.7
- Terminal-Bench 2.0Gemini 3 Flash Preview by 1.246.447.6+1.2
- HMMT Feb. 2025tie97.597.5tie
- HMMT Nov. 2025tie93.393.3tie
Questions people ask
Which is better, DeepSeek-V3.2-Speciale or Gemini 3 Flash Preview?
Gemini 3 Flash Preview wins five of the six areas where both have results: coding, agents, reasoning, long documents and following instructions. DeepSeek-V3.2-Speciale wins none, but costs 79% less. They are level on facts.
Which is better for coding?
Gemini 3 Flash Preview. It wins 3 of the 3 coding tests both models report; DeepSeek-V3.2-Speciale wins none.
Which is cheaper?
DeepSeek-V3.2-Speciale costs $0.29 per million input tokens and $0.43 per million output tokens; Gemini 3 Flash Preview costs $0.50 and $3.00. That makes DeepSeek-V3.2-Speciale about 79% 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 12 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.