DeepSeek V4 Flash vs Gemini 3.5 Flash
DeepSeek V4 Flash 
Gemini 3.5 Flash Verdict
- Gemini 3.5 Flash (medium) leads on quality: 64.2 vs 62.1.
- DeepSeek V4 Flash (max) is stronger in composite, instruction following, long context, reasoning.
- Gemini 3.5 Flash (medium) is stronger in agents & tools, coding, knowledge, maths, human preference, multimodal.
- DeepSeek V4 Flash (max) is 13× cheaper ($0.262 vs $3.38 per 1M blended).
| Metric | DeepSeek V4 Flash (max) | Gemini 3.5 Flash (medium) |
|---|---|---|
| BenchLeader Index | 62.1 | 64.2 |
| Agents & tools score | 67.0 | 67.9 |
| Coding score | 49.1 | 58.7 |
| Composite score | 72.8 | 71.6 |
| Instruction following score | 76.7 | 72.7 |
| Knowledge score | 57.4 | 74.1 |
| Long context score | 66.3 | 63.5 |
| Maths score | 57.3 | 67.2 |
| Reasoning score | 73.9 | 64.2 |
| Human preference score | – | 67.4 |
| Multimodal score | – | 67.5 |
| Blended price $/M | $0.262 | $3.38 |
| Output speed | 212 tok/s | 201 tok/s |
| Time to first answer | 10.6 s | 13.9 s |
| Context window | 1M | 1.0M |
| SciCode | 44.9% | – |
| WeirdML | 45.6% | – |
| LMArena Text | – | 1474 |
| LMArena Hard Prompts | – | 1493 |
| LMArena Coding | – | 1502 |
| LMArena WebDev | – | 1492 |
| LMArena Vision | – | 1307 |
| AA Intelligence Index | 34.5 | 33.6 |
| IFBench | 79.2% | 74.6% |
| AA-LCR | 79.7% | 74.3% |
| MMMU-Pro | – | 83.9% |
| AA-Omniscience | -14.3 | 20.8 |
| Terminal-Bench Hard | 35.6% | 39.4% |
| GPQA Diamond (AA) | 90.8% | 92.1% |
| Humanity's Last Exam (AA) | 38.5% | 41.3% |
| SciCode (AA) | 50.4% | – |
| τ²-Bench Telecom (AA) | 95.0% | 95.6% |
| AIME 2026 | 95.8% | – |
| HMMT February 2026 | 93.9% | – |
| MathArena Apex | 27.1% | – |
| CritPt | 16.6% | 10.9% |
| GDPval (AA) | 47.1% | – |
| τ²-Bench Banking (AA) | 39.4% | – |
| ITBench SRE (AA) | 31.5% | – |
| Analyst Agent (AA) | 25.0% | – |
| LMArena Maths | – | 1478 |
| LMArena Creative Writing | – | 1466 |
| LMArena Instruction Following | – | 1463 |
| LMArena Multi-turn | – | 1476 |
| LMArena Longer Queries | – | 1478 |
| LMArena Document | – | 1461 |
| Surface Evolver Bench | – | 58.1% |
| DeepSWE | – | 37.4% |
| LMCA | 35.9% | – |
| DTBench | 86.4% | – |
| GDP.pdf | – | 14.0% |
Data as of 2026-09-19. Best configuration of each model; every score links to its source on the model pages.
DeepSeek V4 Flash vs Gemini 3.5 Flash: questions
- Is DeepSeek V4 Flash better than Gemini 3.5 Flash?
- Gemini 3.5 Flash (medium) leads on quality: 64.2 vs 62.1. The BenchLeader Index combines every independent quality benchmark; Gemini 3.5 Flash (medium) is ahead overall as of 2026-09-19, but check the category scores for your use.
- Is DeepSeek V4 Flash better than Gemini 3.5 Flash for coding?
- Gemini 3.5 Flash scores higher in coding (59 vs 49 on the category index, where 50 is average).
- Is DeepSeek V4 Flash better than Gemini 3.5 Flash for agentic tasks?
- Gemini 3.5 Flash scores higher in agentic tasks (68 vs 67 on the category index, where 50 is average).
- Which is cheaper, DeepSeek V4 Flash or Gemini 3.5 Flash?
- DeepSeek V4 Flash is cheaper: $0.262 against $3.38 per million tokens, blended at three input tokens per output token.
- Which is faster, DeepSeek V4 Flash or Gemini 3.5 Flash?
- DeepSeek V4 Flash streams faster: 212 against 201 output tokens per second.
- Which has the larger context window?
- Gemini 3.5 Flash accepts more context: 1.0M against 1M tokens.