Gemini 3.8 Flash vs Gemini 3 Flash
Gemini 3.8 Flash Verdict
- Gemini 3.8 Flash (medium) leads on quality: 63.6 vs 61.6.
- Gemini 3.8 Flash (medium) is stronger in coding, composite, knowledge, long context, multimodal.
- Gemini 3 Flash (thinking) is stronger in agents & tools, instruction following.
- Gemini 3 Flash (thinking) is 1.3× cheaper ($1.13 vs $1.50 per 1M blended).
- Gemini 3 Flash (thinking) streams 2.0× faster (180 vs 90 tokens per second).
| Metric | Gemini 3.8 Flash (medium) | Gemini 3 Flash (thinking) |
|---|---|---|
| BenchLeader Index | 63.6 | 61.6 |
| Coding score | 63.8 | – |
| Composite score | 79.6 | 62.3 |
| Knowledge score | 77.7 | 69.0 |
| Long context score | 68.6 | 65.5 |
| Multimodal score | 68.8 | 64.4 |
| Agents & tools score | – | 67.4 |
| Instruction following score | – | 75.4 |
| Blended price $/M | $1.50 | $1.13 |
| Output speed | 90 tok/s | 180 tok/s |
| Time to first answer | 11.7 s | 6.1 s |
| Context window | 1M | 1M |
| SciCode | 54.4% | – |
| AA Intelligence Index | 40 | 26.3 |
| IFBench | – | 78.0% |
| AA-LCR | 84.0% | 78.0% |
| MMMU-Pro | 84.2% | 79.9% |
| AA-Omniscience | 28.6 | 10.1 |
| Terminal-Bench Hard | – | 38.6% |
| GPQA Diamond (AA) | 93.5% | 89.8% |
| Humanity's Last Exam (AA) | 42.1% | 36.6% |
| SciCode (AA) | 55.1% | – |
| τ²-Bench Telecom (AA) | – | 80.4% |
Data as of 2026-09-10. Best configuration of each model; every score links to its source on the model pages.
Gemini 3.8 Flash vs Gemini 3 Flash: questions
- Is Gemini 3.8 Flash better than Gemini 3 Flash?
- Gemini 3.8 Flash (medium) leads on quality: 63.6 vs 61.6. The BenchLeader Index combines every independent quality benchmark; Gemini 3.8 Flash (medium) is ahead overall as of 2026-09-10, but check the category scores for your use.
- Which is cheaper, Gemini 3.8 Flash or Gemini 3 Flash?
- Gemini 3 Flash is cheaper: $1.13 against $1.50 per million tokens, blended at three input tokens per output token.
- Which is faster, Gemini 3.8 Flash or Gemini 3 Flash?
- Gemini 3 Flash streams faster: 180 against 90 output tokens per second.
- Which has the larger context window?
- Both accept 1M tokens of context.