Model comparison
Gemini 3.5 Flash vs Qwen3.8 27B
Gemini 3.5 Flash is the stronger model overall, scoring 54.1 to 45.3 on the Noometry Index. Qwen3.8 27B costs 3.0× less per token, which makes it the better buy when Gemini 3.5 Flash's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
Google
Alibaba (Qwen)Summary
- They share 27 benchmarks with published results for both. Gemini 3.5 Flash scores higher in 7 categories and Qwen3.8 27B in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.5 Flash leads 62.7 to 31.5.
- The biggest single-benchmark swing is NYT Connections (extended): 92.6% for Gemini 3.5 Flash and 54.5% for Qwen3.8 27B.
- Qwen3.8 27B is cheaper at $0.99 / $1.49 per million input/output tokens, against $1.50 / $9 for Gemini 3.5 Flash.
- Gemini 3.5 Flash accepts more context: 1.05M tokens versus 262K.
- Qwen3.8 27B has downloadable open weights; the other is API-only.
Side by side
| Gemini 3.5 Flash | Qwen3.8 27B | |
|---|---|---|
| Provider | Alibaba (Qwen) | |
| Noometry Index | 54.1 | 45.3 |
| Released | 2026-05-19 | 2026-08-14 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 262K |
| Max output | 66K | 33K |
| Input $ / M tokens | $1.50 | $0.99 |
| Output $ / M tokens | $9 | $1.49 |
| Results tracked | 54 | 28 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 27B leads
Gemini 3.5 Flash: 49.1 (#46), Qwen3.8 27B: 50.3 (#42)
| Benchmark | Gemini 3.5 Flash | Qwen3.8 27B |
|---|---|---|
| LMArena WebDev | 1499 | 1592 |
| SciCode | 53.1% | 46.6% |
| LMArena Coding | 1492 | 1487 |
| SWE-bench Verified | 79.3% | — |
| DeepSWE | 37.4% | — |
| WeirdML | 62.6% | — |
| ALE-Bench | 911.02 | — |
Agentic & Tool Use Not comparable
Gemini 3.5 Flash: 24.8 (#107), Qwen3.8 27B: —
| Benchmark | Gemini 3.5 Flash | Qwen3.8 27B |
|---|---|---|
| APEX-Agents | 27.5% | — |
| GBAEval | 6.7% | — |
| GDP.pdf | 14% | — |
| Vending-Bench 2 | 5,396 | — |
Reasoning Gemini 3.5 Flash leads
Gemini 3.5 Flash: 62.7 (#16), Qwen3.8 27B: 31.5 (#71)
| Benchmark | Gemini 3.5 Flash | Qwen3.8 27B |
|---|---|---|
| NYT Connections (extended) | 92.6% | 54.5% |
| CritPt | 13.1% | 5.4% |
| LMArena Hard Prompts | 1488 | 1459 |
| DTBench | 94.7% | 88% |
| LMCA | 47.1% | 41.4% |
| Surface Evolver Bench | 58.1% | 45% |
| Epoch Capabilities Index | 154.55 | 149.38 |
| ARC-AGI-2 | 72.1% | — |
| SimpleBench | 76.7% | — |
| ARC-AGI-1 | 92.5% | — |
| Chess Puzzles | 50% | — |
| EnigmaEval | 25.4% | — |
| EBR-Bench | 4.8% | — |
| Mystery Game Puzzles | 32% | — |
| ForecastBench | 59 | — |
Math Gemini 3.5 Flash leads
Gemini 3.5 Flash: 60.5 (#30), Qwen3.8 27B: 35.7 (#139)
| Benchmark | Gemini 3.5 Flash | Qwen3.8 27B |
|---|---|---|
| ProofBench | 31% | 16% |
| LMArena Math | 1502 | 1453 |
| FrontierMath (Tiers 1-3) | 62.8% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.3% | — |
| OTIS Mock AIME 2024-2025 | 95.6% | — |
| FrontierMath (Feb 2025 set) | 39% | — |
| FrontierMath Tier 4 (v1) | 14.6% | — |
Knowledge Gemini 3.5 Flash leads
Gemini 3.5 Flash: 66.3 (#10), Qwen3.8 27B: 40.9 (#105)
| Benchmark | Gemini 3.5 Flash | Qwen3.8 27B |
|---|---|---|
| LMArena Expert | 1493 | 1482 |
| GPQA Diamond | 92.8% | — |
| SimpleQA Verified | 66.2% | — |
Multimodal Gemini 3.5 Flash leads
Gemini 3.5 Flash: 45.7 (#13), Qwen3.8 27B: 40.5 (#36)
| Benchmark | Gemini 3.5 Flash | Qwen3.8 27B |
|---|---|---|
| LMArena Vision | 1311 | 1270 |
| Blueprint-Bench 2 | 33.6% | — |
| LMArena Document | 1463 | — |
Multilingual Gemini 3.5 Flash leads
Gemini 3.5 Flash: 57.1 (#12), Qwen3.8 27B: 54.0 (#54)
| Benchmark | Gemini 3.5 Flash | Qwen3.8 27B |
|---|---|---|
| LMArena Non-English | 1475 | 1431 |
| LMArena Chinese | 1525 | 1504 |
| LMArena French | 1490 | 1468 |
| LMArena German | 1500 | 1435 |
| LMArena Japanese | 1481 | 1392 |
| LMArena Korean | 1453 | 1389 |
| LMArena Russian | 1491 | 1421 |
| LMArena Spanish | 1478 | 1444 |
Instruction Following Gemini 3.5 Flash leads
Gemini 3.5 Flash: 77.2 (#29), Qwen3.8 27B: 75.8 (#54)
| Benchmark | Gemini 3.5 Flash | Qwen3.8 27B |
|---|---|---|
| LMArena Instruction Following | 1466 | 1436 |
Long Context Gemini 3.5 Flash leads
Gemini 3.5 Flash: 45.3 (#39), Qwen3.8 27B: 44.3 (#66)
| Benchmark | Gemini 3.5 Flash | Qwen3.8 27B |
|---|---|---|
| LMArena Longer Query | 1481 | 1448 |
Writing & Preference Too close to call
Gemini 3.5 Flash: 65.5 (#46), Qwen3.8 27B: 65.9 (#41)
| Benchmark | Gemini 3.5 Flash | Qwen3.8 27B |
|---|---|---|
| LMArena Text | 1481 | 1441 |
| LMArena Creative Writing | 1467 | 1382 |
| LMArena Multi-Turn | 1482 | 1443 |
| EQ-Bench Creative Writing | — | 1671 |
| EQ-Bench 4 | 1087 | — |
Frequently asked questions
Is Gemini 3.5 Flash better than Qwen3.8 27B?
Gemini 3.5 Flash is the stronger model overall, scoring 54.1 to 45.3 on the Noometry Index. Qwen3.8 27B costs 3.0× less per token, which makes it the better buy when Gemini 3.5 Flash's lead doesn't matter for your workload.
Which is cheaper, Gemini 3.5 Flash or Qwen3.8 27B?
Qwen3.8 27B is cheaper. It lists at $0.99 per million input tokens and $1.49 per million output tokens; Gemini 3.5 Flash lists at $1.50 and $9.
Is Gemini 3.5 Flash or Qwen3.8 27B better for coding?
Qwen3.8 27B scores higher on coding benchmarks: 50.3 versus 49.1 in the Noometry coding category.
Which has the bigger context window?
Gemini 3.5 Flash does, with 1.05M tokens against 262K.
How many benchmarks do Gemini 3.5 Flash and Qwen3.8 27B share?
27 benchmarks have published results for both models. Gemini 3.5 Flash has 54 scored results on Noometry and Qwen3.8 27B has 28.