Model comparison
Gemini 3.1 Flash Lite vs Qwen2.5 72B Instruct
Gemini 3.1 Flash Lite is the stronger model overall, scoring 40.8 to 31.9 on the Noometry Index.
Last verified . 24 shared benchmarks.
Google
Alibaba (Qwen)Summary
- They share 24 benchmarks with published results for both. Gemini 3.1 Flash Lite scores higher in 9 categories and Qwen2.5 72B Instruct in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemini 3.1 Flash Lite leads 40.7 to 19.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 80% for Gemini 3.1 Flash Lite and 8.1% for Qwen2.5 72B Instruct.
- Gemini 3.1 Flash Lite is cheaper at $0.25 / $1.50 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
- Gemini 3.1 Flash Lite accepts more context: 1.05M tokens versus 131K.
- Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.
Side by side
| Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct | |
|---|---|---|
| Provider | Alibaba (Qwen) | |
| Noometry Index | 40.8 | 31.9 |
| Released | 2026-03-03 | 2024-09 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 131K |
| Max output | 66K | 8K |
| Input $ / M tokens | $0.25 | $1.40 |
| Output $ / M tokens | $1.50 | $5.60 |
| Results tracked | 38 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3.1 Flash Lite leads
Gemini 3.1 Flash Lite: 37.8 (#188), Qwen2.5 72B Instruct: 33.2 (#260)
| Benchmark | Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct |
|---|---|---|
| WeirdML | 52.2% | 16% |
| LMArena Coding | 1400 | 1292 |
| LMArena WebDev | 1256 | — |
| SciCode | 41.9% | — |
| BigCodeBench Instruct | — | 45.8% |
| BigCodeBench Complete | — | 55.9% |
| ALE-Bench | 797.73 | — |
Agentic & Tool Use Gemini 3.1 Flash Lite leads
Gemini 3.1 Flash Lite: 30.2 (#79), Qwen2.5 72B Instruct: 22.1 (#133)
| Benchmark | Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct |
|---|---|---|
| TheAgentCompany | — | 5.7% |
| DeepResearch Bench | 37.3% | — |
| BALROG | — | 16.2% |
| METR Time Horizons | — | 35.8% |
Reasoning Too close to call
Gemini 3.1 Flash Lite: 22.9 (#186), Qwen2.5 72B Instruct: 22.3 (#199)
| Benchmark | Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1407 | 1271 |
| DTBench | 76.8% | 62.9% |
| LMCA | 35% | 13.4% |
| Epoch Capabilities Index | 144.47 | 129 |
| ForecastBench | 54.4 | 57.5 |
| Kagi LLM Benchmark | 67.2% | — |
| NYT Connections (extended) | 8.2% | — |
| CritPt | 1.1% | — |
| Chess Puzzles | 25% | — |
| EnigmaEval | 3% | — |
| Thematic Generalization | 63.3% | — |
| BIG-Bench Hard | — | 79.8% |
| HellaSwag | — | 84.8% |
| PIQA | — | 82.6% |
| WinoGrande | — | 82.3% |
Math Gemini 3.1 Flash Lite leads
Gemini 3.1 Flash Lite: 40.7 (#90), Qwen2.5 72B Instruct: 19.3 (#287)
| Benchmark | Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 80% | 8.1% |
| LMArena Math | 1428 | 1283 |
| FrontierMath (Tiers 1-3) | 27.7% | — |
| Omni-MATH | — | 33% |
| MATH Level 5 | — | 63.2% |
Knowledge Gemini 3.1 Flash Lite leads
Gemini 3.1 Flash Lite: 41.9 (#104), Qwen2.5 72B Instruct: 27.0 (#253)
| Benchmark | Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct |
|---|---|---|
| GPQA Diamond | 81.8% | 49.1% |
| LMArena Expert | 1398 | 1245 |
| Humanity's Last Exam | 8.6% | — |
| MMLU-Pro | — | 63.1% |
| Confabulations | — | 19.1% |
| Vectara Hallucination Rate | 8.2% | — |
| GPQA (HELM) | — | 42.6% |
| ARC (AI2) Challenge | — | 94.5% |
| MMLU | — | 85.3% |
| TriviaQA | — | 71.9% |
Multimodal Not comparable
Gemini 3.1 Flash Lite: 39.4 (#60), Qwen2.5 72B Instruct: —
| Benchmark | Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Vision | 1240 | — |
Multilingual Gemini 3.1 Flash Lite leads
Gemini 3.1 Flash Lite: 52.3 (#86), Qwen2.5 72B Instruct: 41.0 (#213)
| Benchmark | Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Non-English | 1411 | 1252 |
| LMArena Chinese | 1461 | 1272 |
| LMArena French | 1424 | 1280 |
| LMArena German | 1429 | 1234 |
| LMArena Japanese | 1413 | 1180 |
| LMArena Korean | 1392 | 1188 |
| LMArena Russian | 1420 | 1264 |
| LMArena Spanish | 1421 | 1256 |
Instruction Following Gemini 3.1 Flash Lite leads
Gemini 3.1 Flash Lite: 72.7 (#131), Qwen2.5 72B Instruct: 65.5 (#221)
| Benchmark | Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Instruction Following | 1377 | 1254 |
| IFEval | — | 80.6% |
Long Context Gemini 3.1 Flash Lite leads
Gemini 3.1 Flash Lite: 42.5 (#122), Qwen2.5 72B Instruct: 38.9 (#188)
| Benchmark | Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Longer Query | 1394 | 1282 |
Writing & Preference Gemini 3.1 Flash Lite leads
Gemini 3.1 Flash Lite: 60.9 (#94), Qwen2.5 72B Instruct: 46.7 (#215)
| Benchmark | Gemini 3.1 Flash Lite | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Text | 1416 | 1269 |
| LMArena Creative Writing | 1401 | 1221 |
| LMArena Multi-Turn | 1417 | 1272 |
| WildBench | — | 80.2% |
Frequently asked questions
Is Gemini 3.1 Flash Lite better than Qwen2.5 72B Instruct?
Gemini 3.1 Flash Lite is the stronger model overall, scoring 40.8 to 31.9 on the Noometry Index.
Which is cheaper, Gemini 3.1 Flash Lite or Qwen2.5 72B Instruct?
Gemini 3.1 Flash Lite is cheaper. It lists at $0.25 per million input tokens and $1.50 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.
Is Gemini 3.1 Flash Lite or Qwen2.5 72B Instruct better for coding?
Gemini 3.1 Flash Lite scores higher on coding benchmarks: 37.8 versus 33.2 in the Noometry coding category.
Which has the bigger context window?
Gemini 3.1 Flash Lite does, with 1.05M tokens against 131K.
How many benchmarks do Gemini 3.1 Flash Lite and Qwen2.5 72B Instruct share?
24 benchmarks have published results for both models. Gemini 3.1 Flash Lite has 38 scored results on Noometry and Qwen2.5 72B Instruct has 43.