← comparisons / gemini-3.7-flash-vs-qwen3-8-27b
Gemini 3.7 Flash vs Qwen3.8 27B
01Head to head
| Gemini 3.7 Flash | Qwen3.8 27B | |
|---|---|---|
| Creator | Alibaba Qwen | |
| Released | 2026-08-13 | 2026-08-14 |
| Context | 1M | 262K |
| Max output | 66K | 33K |
| Input $/M | — | — |
| Output $/M | — | — |
| Knowledge cutoff | 2026-03 | — |
| Modalities | text + image + video + audio → text | text + image + video → text |
| Capabilities | tool-use, extended-thinking, structured-output | — |
| Providers | Gemini API (AI Studio) |
02Benchmarks · vendor-reported unless verified
| Gemini 3.7 Flash | Qwen3.8 27B | |
|---|---|---|
| AutomationBench (accuracy) | 30.4 | — |
| DeepSWE | 65.3% | — |
| FrontierCode (score) | 43.6 | — |
| GDM-MRCR (accuracy) | 97 | — |
| GDP.pdf (accuracy) | 34 | — |
| Terminal-Bench (accuracy) | 85.8 | — |