REVIEWS / COMPARE
DeepSeek R1 vs Gemini 3.6 Flash Family
It's close — DeepSeek R1 (8.8) and Gemini 3.6 Flash Family (8.8) score nearly the same across real user voices. Pick on the specifics below.
| DeepSeek R1 | Gemini 3.6 Flash Family | |
|---|---|---|
| RATING | 8.8 / 10 | 8.8 / 10 |
| CONFIDENCE | high | low |
| USER VOICES | 429 | 13 |
| POSITIVE | 70% | 75% |
| PRICE | — | — |
| RECOMMENDED | ✓ Yes | ✓ Yes |
DeepSeek R1
★ STRENGTHS
- Superior performance on math and coding benchmarks (MATH-500, Codeforces)
- Highly efficient training architecture ('Green energy of AI')
- Distilled smaller models (8B/70B) offer strong local performance
✗ WATCH OUT
- Significant latency; local inference can be 20x slower than proprietary alternatives
- Inconsistent censorship behaviors between cloud and local deployments
- Struggles with nuance, humor, and creative writing tasks
Gemini 3.6 Flash Family
★ STRENGTHS
- Exceptional multimodal capabilities (audio, images, interactive elements)
- High-quality UI/UX and web design asset generation
- Strong token efficiency compared to competing models
✗ WATCH OUT
- Concerns about latency and reliability in long multi-step agent runs
- Users report the 'normal model' may have been nerfed
- Balancing safety guardrails without excessive refusals remains a questioned area
HOW THIS COMPARISON WAS BUILT
Both verdicts are synthesised from real user voices on Reddit, YouTube, Hacker News, ProductHunt, Trustpilot and editorial sources — 429 for DeepSeek R1, 13 for Gemini 3.6 Flash Family. No paid placement; ratings come from raw sentiment, not our opinion. Read the methodology.