Model comparison
Qwen2.5 72B Instruct vs Trinity Large Thinking
Trinity Large Thinking is the stronger model overall, scoring 38.6 to 31.9 on the Noometry Index.
Last verified . 17 shared benchmarks.
Alibaba (Qwen)
Arcee AISummary
- They share 17 benchmarks with published results for both. Qwen2.5 72B Instruct scores higher in 1 category and Trinity Large Thinking in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Trinity Large Thinking leads 37.6 to 19.3.
- Trinity Large Thinking is cheaper at $0.25 / $0.80 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
- Trinity Large Thinking accepts more context: 262K tokens versus 131K.
Side by side
| Qwen2.5 72B Instruct | Trinity Large Thinking | |
|---|---|---|
| Provider | Alibaba (Qwen) | Arcee AI |
| Noometry Index | 31.9 | 38.6 |
| Released | 2024-09 | 2026-04-01 |
| Weights | Open | Open |
| Context window | 131K | 262K |
| Max output | 8K | 80K |
| Input $ / M tokens | $1.40 | $0.25 |
| Output $ / M tokens | $5.60 | $0.80 |
| Results tracked | 43 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Qwen2.5 72B Instruct: 33.2 (#260), Trinity Large Thinking: 34.1 (#244)
| Benchmark | Qwen2.5 72B Instruct | Trinity Large Thinking |
|---|---|---|
| LMArena Coding | 1292 | 1381 |
| LMArena WebDev | — | 1238 |
| SciCode | — | 36.1% |
| WeirdML | 16% | — |
| BigCodeBench Instruct | 45.8% | — |
| BigCodeBench Complete | 55.9% | — |
Agentic & Tool Use Not comparable
Qwen2.5 72B Instruct: 22.1 (#133), Trinity Large Thinking: —
| Benchmark | Qwen2.5 72B Instruct | Trinity Large Thinking |
|---|---|---|
| TheAgentCompany | 5.7% | — |
| BALROG | 16.2% | — |
| METR Time Horizons | 35.8% | — |
Reasoning Qwen2.5 72B Instruct leads
Qwen2.5 72B Instruct: 22.3 (#199), Trinity Large Thinking: 16.9 (#298)
| Benchmark | Qwen2.5 72B Instruct | Trinity Large Thinking |
|---|---|---|
| LMArena Hard Prompts | 1271 | 1350 |
| NYT Connections (extended) | — | 16.5% |
| CritPt | — | 0.9% |
| Thematic Generalization | — | 41.6% |
| DTBench | 62.9% | — |
| LMCA | 13.4% | — |
| Surface Evolver Bench | — | 15.6% |
| BIG-Bench Hard | 79.8% | — |
| Epoch Capabilities Index | 129 | — |
| ForecastBench | 57.5 | — |
| HellaSwag | 84.8% | — |
| PIQA | 82.6% | — |
| WinoGrande | 82.3% | — |
Math Trinity Large Thinking leads
Qwen2.5 72B Instruct: 19.3 (#287), Trinity Large Thinking: 37.6 (#149)
| Benchmark | Qwen2.5 72B Instruct | Trinity Large Thinking |
|---|---|---|
| LMArena Math | 1283 | 1366 |
| OTIS Mock AIME 2024-2025 | 8.1% | — |
| Omni-MATH | 33% | — |
| MATH Level 5 | 63.2% | — |
Knowledge Trinity Large Thinking leads
Qwen2.5 72B Instruct: 27.0 (#253), Trinity Large Thinking: 40.9 (#113)
| Benchmark | Qwen2.5 72B Instruct | Trinity Large Thinking |
|---|---|---|
| LMArena Expert | 1245 | 1360 |
| GPQA Diamond | 49.1% | — |
| MMLU-Pro | 63.1% | — |
| Confabulations | 19.1% | — |
| Vectara Hallucination Rate | — | 6.9% |
| GPQA (HELM) | 42.6% | — |
| ARC (AI2) Challenge | 94.5% | — |
| MMLU | 85.3% | — |
| TriviaQA | 71.9% | — |
Multilingual Trinity Large Thinking leads
Qwen2.5 72B Instruct: 41.0 (#213), Trinity Large Thinking: 46.2 (#160)
| Benchmark | Qwen2.5 72B Instruct | Trinity Large Thinking |
|---|---|---|
| LMArena Non-English | 1252 | 1325 |
| LMArena Chinese | 1272 | 1373 |
| LMArena French | 1280 | 1374 |
| LMArena German | 1234 | 1356 |
| LMArena Japanese | 1180 | 1311 |
| LMArena Korean | 1188 | 1306 |
| LMArena Russian | 1264 | 1337 |
| LMArena Spanish | 1256 | 1357 |
Instruction Following Trinity Large Thinking leads
Qwen2.5 72B Instruct: 65.5 (#221), Trinity Large Thinking: 70.5 (#162)
| Benchmark | Qwen2.5 72B Instruct | Trinity Large Thinking |
|---|---|---|
| LMArena Instruction Following | 1254 | 1334 |
| IFEval | 80.6% | — |
Long Context Trinity Large Thinking leads
Qwen2.5 72B Instruct: 38.9 (#188), Trinity Large Thinking: 41.3 (#144)
| Benchmark | Qwen2.5 72B Instruct | Trinity Large Thinking |
|---|---|---|
| LMArena Longer Query | 1282 | 1355 |
Writing & Preference Trinity Large Thinking leads
Qwen2.5 72B Instruct: 46.7 (#215), Trinity Large Thinking: 53.8 (#158)
| Benchmark | Qwen2.5 72B Instruct | Trinity Large Thinking |
|---|---|---|
| LMArena Text | 1269 | 1340 |
| LMArena Creative Writing | 1221 | 1320 |
| LMArena Multi-Turn | 1272 | 1342 |
| WildBench | 80.2% | — |
Frequently asked questions
Is Qwen2.5 72B Instruct better than Trinity Large Thinking?
Trinity Large Thinking is the stronger model overall, scoring 38.6 to 31.9 on the Noometry Index.
Which is cheaper, Qwen2.5 72B Instruct or Trinity Large Thinking?
Trinity Large Thinking is cheaper. It lists at $0.25 per million input tokens and $0.80 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.
Is Qwen2.5 72B Instruct or Trinity Large Thinking better for coding?
They score almost the same on coding (33.2 vs 34.1); test both on your own repository before choosing.
Which has the bigger context window?
Trinity Large Thinking does, with 262K tokens against 131K.
How many benchmarks do Qwen2.5 72B Instruct and Trinity Large Thinking share?
17 benchmarks have published results for both models. Qwen2.5 72B Instruct has 43 scored results on Noometry and Trinity Large Thinking has 24.