Model comparison
Gemini 2.5 Flash-Lite vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 37.0 on the Noometry Index. Gemini 2.5 Flash-Lite costs 17× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Last verified . 21 shared benchmarks.
Google
Alibaba (Qwen)Summary
- They share 21 benchmarks with published results for both. Gemini 2.5 Flash-Lite scores higher in 0 categories and Qwen3.8 Max in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 38.0.
- The biggest single-benchmark swing is DTBench: 62.8% for Gemini 2.5 Flash-Lite and 92% for Qwen3.8 Max.
- Gemini 2.5 Flash-Lite is cheaper at $0.10 / $0.40 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
- Gemini 2.5 Flash-Lite accepts more context: 1.05M tokens versus 1M.
Side by side
| Gemini 2.5 Flash-Lite | Qwen3.8 Max | |
|---|---|---|
| Provider | Alibaba (Qwen) | |
| Noometry Index | 37.0 | 56.8 |
| Released | 2025-06-17 | 2026-08-02 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1M |
| Max output | 66K | 131K |
| Input $ / M tokens | $0.10 | $2 |
| Output $ / M tokens | $0.40 | $6 |
| Results tracked | 33 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
Gemini 2.5 Flash-Lite: 38.5 (#173), Qwen3.8 Max: 53.5 (#29)
| Benchmark | Gemini 2.5 Flash-Lite | Qwen3.8 Max |
|---|---|---|
| LMArena Coding | 1373 | 1502 |
| DeepSWE | — | 57.5% |
| LMArena WebDev | — | 1674 |
| FrontierSWE | — | 17.8% |
| SciCode | — | 53.2% |
| WeirdML | 35.2% | — |
| ALE-Bench | 325.9 | — |
Agentic & Tool Use Qwen3.8 Max leads
Gemini 2.5 Flash-Lite: 28.0 (#96), Qwen3.8 Max: 45.4 (#14)
| Benchmark | Gemini 2.5 Flash-Lite | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | — | 63.3% |
| Berkeley Function Calling Leaderboard | 36.9% | — |
| τ²-bench Banking | — | 55.1% |
| GDP.pdf | — | 23.2% |
Reasoning Qwen3.8 Max leads
Gemini 2.5 Flash-Lite: 22.2 (#205), Qwen3.8 Max: 54.4 (#26)
| Benchmark | Gemini 2.5 Flash-Lite | Qwen3.8 Max |
|---|---|---|
| LMArena Hard Prompts | 1377 | 1496 |
| DTBench | 62.8% | 92% |
| LMCA | 18.1% | 46.2% |
| Epoch Capabilities Index | 133.94 | 156.41 |
| Kagi LLM Benchmark | 40.5% | — |
| NYT Connections (extended) | — | 88.3% |
| CritPt | — | 20% |
| Chess Puzzles | — | 40% |
| Mystery Game Puzzles | — | 38% |
Math Qwen3.8 Max leads
Gemini 2.5 Flash-Lite: 38.0 (#144), Qwen3.8 Max: 73.2 (#20)
| Benchmark | Gemini 2.5 Flash-Lite | Qwen3.8 Max |
|---|---|---|
| LMArena Math | 1373 | 1499 |
| FrontierMath (Tiers 1-3) | — | 74.7% |
| FrontierMath Tier 4 | — | 46.3% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| ProofBench | — | 58% |
| Omni-MATH | 48% | — |
Knowledge Qwen3.8 Max leads
Gemini 2.5 Flash-Lite: 32.5 (#210), Qwen3.8 Max: 61.7 (#27)
| Benchmark | Gemini 2.5 Flash-Lite | Qwen3.8 Max |
|---|---|---|
| LMArena Expert | 1373 | 1507 |
| GPQA Diamond | — | 92.7% |
| SimpleQA Verified | — | 47.3% |
| MMLU-Pro | 53.7% | — |
| Vectara Hallucination Rate | 3.3% | — |
| GPQA (HELM) | 30.9% | — |
Multimodal Qwen3.8 Max leads
Gemini 2.5 Flash-Lite: 29.1 (#114), Qwen3.8 Max: 37.2 (#75)
| Benchmark | Gemini 2.5 Flash-Lite | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | 1198 | 1314 |
| VPCT | 30% | — |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
Gemini 2.5 Flash-Lite: 49.3 (#134), Qwen3.8 Max: 56.7 (#18)
| Benchmark | Gemini 2.5 Flash-Lite | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1369 | 1472 |
| LMArena Chinese | 1404 | 1538 |
| LMArena French | 1388 | 1503 |
| LMArena German | 1389 | 1483 |
| LMArena Japanese | 1359 | 1467 |
| LMArena Korean | 1360 | 1461 |
| LMArena Russian | 1373 | 1481 |
| LMArena Spanish | 1396 | 1492 |
Instruction Following Qwen3.8 Max leads
Gemini 2.5 Flash-Lite: 70.0 (#168), Qwen3.8 Max: 77.6 (#17)
| Benchmark | Gemini 2.5 Flash-Lite | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1367 | 1479 |
| IFEval | 81% | — |
Long Context Qwen3.8 Max leads
Gemini 2.5 Flash-Lite: 33.3 (#262), Qwen3.8 Max: 45.6 (#31)
| Benchmark | Gemini 2.5 Flash-Lite | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1373 | 1489 |
| Fiction.LiveBench | 47.2% | — |
Writing & Preference Qwen3.8 Max leads
Gemini 2.5 Flash-Lite: 56.8 (#135), Qwen3.8 Max: 67.1 (#30)
| Benchmark | Gemini 2.5 Flash-Lite | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1379 | 1483 |
| LMArena Creative Writing | 1367 | 1479 |
| LMArena Multi-Turn | 1366 | 1489 |
| WildBench | 81.8% | — |
Frequently asked questions
Is Gemini 2.5 Flash-Lite better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 37.0 on the Noometry Index. Gemini 2.5 Flash-Lite costs 17× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Which is cheaper, Gemini 2.5 Flash-Lite or Qwen3.8 Max?
Gemini 2.5 Flash-Lite is cheaper. It lists at $0.10 per million input tokens and $0.40 per million output tokens; Qwen3.8 Max lists at $2 and $6.
Is Gemini 2.5 Flash-Lite or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 38.5 in the Noometry coding category.
Which has the bigger context window?
Gemini 2.5 Flash-Lite does, with 1.05M tokens against 1M.
How many benchmarks do Gemini 2.5 Flash-Lite and Qwen3.8 Max share?
21 benchmarks have published results for both models. Gemini 2.5 Flash-Lite has 33 scored results on Noometry and Qwen3.8 Max has 39.