Model comparison
Qwen3 235B-A22B vs Trinity Large Thinking
Qwen3 235B-A22B is the stronger model overall, scoring 43.2 to 38.1 on the Noometry Index. Trinity Large Thinking costs 3.2× less per token, which makes it the better buy when Qwen3 235B-A22B's lead doesn't matter for your workload.
Last verified . 20 shared benchmarks.
Alibaba (Qwen)
Arcee AISummary
- They share 20 benchmarks with published results for both. Qwen3 235B-A22B scores higher in 7 categories and Trinity Large Thinking in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3 235B-A22B leads 48.5 to 34.9.
- The biggest single-benchmark swing is SciCode: 42.4% for Qwen3 235B-A22B and 36.1% for Trinity Large Thinking.
- Trinity Large Thinking is cheaper at $0.25 / $0.80 per million input/output tokens, against $0.70 / $2.80 for Qwen3 235B-A22B.
- Trinity Large Thinking accepts more context: 262K tokens versus 131K.
Side by side
| Qwen3 235B-A22B | Trinity Large Thinking | |
|---|---|---|
| Provider | Alibaba (Qwen) | Arcee AI |
| Noometry Index | 43.2 | 38.1 |
| Released | 2025-04 | 2026-04-01 |
| Weights | Open | Open |
| Context window | 131K | 262K |
| Max output | 16K | 80K |
| Input $ / M tokens | $0.70 | $0.25 |
| Output $ / M tokens | $2.80 | $0.80 |
| Results tracked | 49 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 235B-A22B leads
Qwen3 235B-A22B: 43.8 (#72), Trinity Large Thinking: 33.8 (#238)
| Benchmark | Qwen3 235B-A22B | Trinity Large Thinking |
|---|---|---|
| SciCode | 42.4% | 36.1% |
| LMArena Coding | 1445 | 1381 |
| Aider Polyglot | 59.6% | — |
| LMArena WebDev | — | 1238 |
| WeirdML | 41% | — |
Agentic & Tool Use Not comparable
Qwen3 235B-A22B: 33.7 (#48), Trinity Large Thinking: —
| Benchmark | Qwen3 235B-A22B | Trinity Large Thinking |
|---|---|---|
| Berkeley Function Calling Leaderboard | 52.1% | — |
| Vending-Bench 2 | -11.34 | — |
Reasoning Too close to call
Qwen3 235B-A22B: 15.8 (#302), Trinity Large Thinking: 16.3 (#298)
| Benchmark | Qwen3 235B-A22B | Trinity Large Thinking |
|---|---|---|
| CritPt | 0% | 0.9% |
| LMArena Hard Prompts | 1433 | 1350 |
| ARC-AGI-2 | 1.3% | — |
| SimpleBench | 31% | — |
| Kagi LLM Benchmark | 69.4% | — |
| NYT Connections (extended) | — | 16.5% |
| ARC-AGI-1 | 11% | — |
| Chess Puzzles | 12% | — |
| Thematic Generalization | — | 41.6% |
| Mystery Game Puzzles | 9% | — |
| DTBench | 80.3% | — |
| LMCA | 29.3% | — |
| Surface Evolver Bench | — | 15.6% |
| Epoch Capabilities Index | 143.87 | — |
| ForecastBench | 59.7 | — |
Math Qwen3 235B-A22B leads
Qwen3 235B-A22B: 48.5 (#51), Trinity Large Thinking: 34.9 (#149)
| Benchmark | Qwen3 235B-A22B | Trinity Large Thinking |
|---|---|---|
| LMArena Math | 1431 | 1366 |
| OTIS Mock AIME 2024-2025 | 86.7% | — |
| Omni-MATH | 71.8% | — |
| MATH Level 5 | 68.9% | — |
| FrontierMath (Feb 2025 set) | 8.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Qwen3 235B-A22B leads
Qwen3 235B-A22B: 49.5 (#67), Trinity Large Thinking: 40.3 (#107)
| Benchmark | Qwen3 235B-A22B | Trinity Large Thinking |
|---|---|---|
| Vectara Hallucination Rate | 9.3% | 6.9% |
| LMArena Expert | 1460 | 1359 |
| GPQA Diamond | 80.1% | — |
| SimpleQA Verified | 40.4% | — |
| MMLU-Pro | 84.4% | — |
| Confabulations | 15.6% | — |
| GPQA (HELM) | 72.7% | — |
Multilingual Qwen3 235B-A22B leads
Qwen3 235B-A22B: 52.5 (#84), Trinity Large Thinking: 46.4 (#155)
| Benchmark | Qwen3 235B-A22B | Trinity Large Thinking |
|---|---|---|
| LMArena Non-English | 1410 | 1325 |
| LMArena Chinese | 1481 | 1372 |
| LMArena French | 1445 | 1374 |
| LMArena German | 1432 | 1356 |
| LMArena Japanese | 1400 | 1312 |
| LMArena Korean | 1391 | 1304 |
| LMArena Russian | 1411 | 1338 |
| LMArena Spanish | 1429 | 1356 |
Instruction Following Qwen3 235B-A22B leads
Qwen3 235B-A22B: 72.8 (#132), Trinity Large Thinking: 70.7 (#157)
| Benchmark | Qwen3 235B-A22B | Trinity Large Thinking |
|---|---|---|
| LMArena Instruction Following | 1408 | 1334 |
| IFEval | 83.5% | — |
Long Context Qwen3 235B-A22B leads
Qwen3 235B-A22B: 46.1 (#25), Trinity Large Thinking: 41.2 (#139)
| Benchmark | Qwen3 235B-A22B | Trinity Large Thinking |
|---|---|---|
| LMArena Longer Query | 1426 | 1355 |
| Fiction.LiveBench | 75% | — |
Writing & Preference Qwen3 235B-A22B leads
Qwen3 235B-A22B: 59.8 (#103), Trinity Large Thinking: 54.0 (#152)
| Benchmark | Qwen3 235B-A22B | Trinity Large Thinking |
|---|---|---|
| LMArena Text | 1419 | 1340 |
| LMArena Creative Writing | 1384 | 1321 |
| LMArena Multi-Turn | 1432 | 1342 |
| Short-Story Creative Writing | 83% | — |
| EQ-Bench Creative Writing | 1366 | — |
| WildBench | 86.6% | — |
Frequently asked questions
Is Qwen3 235B-A22B better than Trinity Large Thinking?
Qwen3 235B-A22B is the stronger model overall, scoring 43.2 to 38.1 on the Noometry Index. Trinity Large Thinking costs 3.2× less per token, which makes it the better buy when Qwen3 235B-A22B's lead doesn't matter for your workload.
Which is cheaper, Qwen3 235B-A22B 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; Qwen3 235B-A22B lists at $0.70 and $2.80.
Is Qwen3 235B-A22B or Trinity Large Thinking better for coding?
Qwen3 235B-A22B scores higher on coding benchmarks: 43.8 versus 33.8 in the Noometry coding category.
Which has the bigger context window?
Trinity Large Thinking does, with 262K tokens against 131K.
How many benchmarks do Qwen3 235B-A22B and Trinity Large Thinking share?
20 benchmarks have published results for both models. Qwen3 235B-A22B has 49 scored results on Noometry and Trinity Large Thinking has 24.