Gemini 3.5 Flash vs GPT-5.4 Pro
Gemini 3.5 Flash 
GPT-5.4 Pro Verdict
- Gemini 3.5 Flash (medium) and GPT-5.4 Pro are level on quality (63.9 vs 64.2).
- Gemini 3.5 Flash (medium) is stronger in agents & tools, coding, composite, human preference, instruction following, knowledge, long context.
- GPT-5.4 Pro is stronger in multimodal, reasoning.
- Gemini 3.5 Flash (medium) is 20× cheaper ($3.38 vs $67.50 per 1M blended).
- Gemini 3.5 Flash (medium) streams 208.9× faster (209 vs 1 tokens per second).
| Metric | Gemini 3.5 Flash (medium) | GPT-5.4 Pro |
|---|---|---|
| BenchLeader Index | 63.9 | 64.2 |
| Agents & tools score | 67.8 | – |
| Coding score | 58.9 | – |
| Composite score | 71.4 | – |
| Human preference score | 67.4 | – |
| Instruction following score | 72.2 | 67.3 |
| Knowledge score | 73.9 | 73.2 |
| Long context score | 63.6 | – |
| Multimodal score | 67.6 | 69.8 |
| Reasoning score | 66.8 | 75.1 |
| Blended price $/M | $3.38 | $67.50 |
| Output speed | 209 tok/s | 1 tok/s |
| Time to first answer | 17.0 s | 6.7 s |
| Context window | 1M | 1.1M |
| Humanity's Last Exam | – | 44.3% |
| SimpleBench | – | 74.1% |
| Epoch Capabilities Index | – | 158.9 |
| LMArena Text | 1476 | – |
| LMArena Hard Prompts | 1493 | – |
| LMArena Coding | 1503 | – |
| LMArena WebDev | 1491 | – |
| LMArena Vision | 1306 | – |
| AA Intelligence Index | 33.6 | – |
| IFBench | 74.6% | – |
| AA-LCR | 74.3% | – |
| MMMU-Pro | 83.9% | – |
| AA-Omniscience | 20.8 | – |
| Terminal-Bench Hard | 39.4% | – |
| GPQA Diamond (AA) | 92.1% | – |
| Humanity's Last Exam (AA) | 41.3% | – |
| τ²-Bench Telecom (AA) | 95.6% | – |
| MultiChallenge | – | 69.2% |
| VISTA | – | 53.9% |
| MultiNRC | – | 62.3% |
| TutorBench | – | 56.6% |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.