Gemini 3.5 Flash-Lite vs Qwen3.5 397B-A17B
| Spec | Gemini 3.5 Flash-Lite | Qwen3.5 397B-A17B |
|---|---|---|
| Respan Index | 29.9 (#93) | 29.5 (#94) |
| Blended price per 1M | $0.85 | $0.387 |
| Context window | 1.0M | 262K |
| Input / output per 1M | $0.3 / $2.5 | $0.172 / $1.03 |
| Max output | 66K | - |
| Input | text, vision, audio | text, vision |
| Released | 2026-07-21 | 2026-02-15 |
Respan Index by category
Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.
| Category | Gemini 3.5 Flash-Lite | Qwen3.5 397B-A17B |
|---|---|---|
| Coding | 6.0 Gemini 3.5 Flash-Lite | 11.2 Qwen3.5 397B-A17B |
| Agents | 44.0 Gemini 3.5 Flash-Lite | 34.3 Qwen3.5 397B-A17B |
| Reasoning | 25.6 Gemini 3.5 Flash-Lite | 30.8 Qwen3.5 397B-A17B |
| Knowledge & language | 32.0 Gemini 3.5 Flash-Lite | 27.6 Qwen3.5 397B-A17B |
| Human preference | 52.0 Gemini 3.5 Flash-Lite | 51.1 Qwen3.5 397B-A17B |
Benchmarks both report
Gemini 3.5 Flash-Lite leads on 7 of the 11 benchmarks both report. Higher is better. Each score links to where it was published.
| Benchmark | Gemini 3.5 Flash-Lite | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Elo | 1455 | 1442 |
| LMArena Vision (LMArena) | 1266 | 1245 |
| LMArena WebDev (LMArena) | 1432 | 1400 |
| SAGE 1.1 (Vals AI) | 79.5% | 85.5% |
| Chess Puzzles (Epoch AI) | 21% | 13% |
| APEX-Agents (Mercor) | 29.3% | 24.9% |
| FrontierMath Tiers 1-3 v2 (Epoch AI) | 26% | 31.2% |
| GPQA Diamond | 83.3% | 88.4% |
| Mystery Game Puzzles (Epoch AI) | 19% | 18% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 71.1% | 88.9% |
| Toolathlon-Verified (HKUST) | 57.1% | 40.7% |
FAQ
Is Gemini 3.5 Flash-Lite better than Qwen3.5 397B-A17B?
Gemini 3.5 Flash-Lite scores 29.9 on the Respan Index (#93), ahead of Qwen3.5 397B-A17B at 29.5 (#94). Gemini 3.5 Flash-Lite leads on 7 of the 11 benchmarks both report.
Which is cheaper, Gemini 3.5 Flash-Lite or Qwen3.5 397B-A17B?
Gemini 3.5 Flash-Lite lists at $0.3 input and $2.5 output per 1M tokens; Qwen3.5 397B-A17B at $0.172 and $1.03. Qwen3.5 397B-A17B is 2.2x cheaper per token at list price.
Which has the larger context window?
Gemini 3.5 Flash-Lite accepts 1.0M tokens and Qwen3.5 397B-A17B 262K tokens.