Model comparison
GPT-4 vs Trinity Large Thinking
Trinity Large Thinking is the stronger model overall, scoring 38.1 to 28.9 on the Noometry Index.
Last verified . 17 shared benchmarks.
Arcee AISummary
- They share 17 benchmarks with published results for both. GPT-4 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 34.9 to 10.1.
- Trinity Large Thinking is cheaper at $0.25 / $0.80 per million input/output tokens, against $30 / $60 for GPT-4.
- Trinity Large Thinking accepts more context: 262K tokens versus 8K.
- Trinity Large Thinking has downloadable open weights; the other is API-only.
Side by side
| GPT-4 | Trinity Large Thinking | |
|---|---|---|
| Provider | OpenAI | Arcee AI |
| Noometry Index | 28.9 | 38.1 |
| Released | 2023-03-14 | 2026-04-01 |
| Weights | Proprietary | Open |
| Context window | 8K | 262K |
| Max output | 8K | 80K |
| Input $ / M tokens | $30 | $0.25 |
| Output $ / M tokens | $60 | $0.80 |
| Results tracked | 38 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Trinity Large Thinking leads
GPT-4: 31.2 (#277), Trinity Large Thinking: 33.8 (#238)
| Benchmark | GPT-4 | Trinity Large Thinking |
|---|---|---|
| LMArena Coding | 1254 | 1381 |
| LMArena WebDev | — | 1238 |
| SciCode | — | 36.1% |
| WeirdML | 12.4% | — |
| BigCodeBench Instruct | 46% | — |
| BigCodeBench Complete | 57.2% | — |
| HumanEval+ | 79.3% | — |
Agentic & Tool Use Not comparable
GPT-4: —, Trinity Large Thinking: —
| Benchmark | GPT-4 | Trinity Large Thinking |
|---|---|---|
| METR Time Horizons | 36.1% | — |
Reasoning Too close to call
GPT-4: 17.3 (#285), Trinity Large Thinking: 16.3 (#298)
| Benchmark | GPT-4 | Trinity Large Thinking |
|---|---|---|
| LMArena Hard Prompts | 1241 | 1350 |
| NYT Connections (extended) | — | 16.5% |
| CritPt | — | 0.9% |
| Chess Puzzles | 4% | — |
| Thematic Generalization | — | 41.6% |
| Mystery Game Puzzles | 12% | — |
| DTBench | 62.7% | — |
| LMCA | 17.1% | — |
| Surface Evolver Bench | — | 15.6% |
| BIG-Bench Hard | 75.1% | — |
| Epoch Capabilities Index | 125.88 | — |
| ForecastBench | 57.8 | — |
| HellaSwag | 95.3% | — |
| WinoGrande | 87.5% | — |
Math Trinity Large Thinking leads
GPT-4: 10.1 (#304), Trinity Large Thinking: 34.9 (#149)
| Benchmark | GPT-4 | Trinity Large Thinking |
|---|---|---|
| LMArena Math | 1269 | 1366 |
| OTIS Mock AIME 2024-2025 | 1.1% | — |
| MATH Level 5 | 23% | — |
| GSM8K | 92% | — |
Knowledge Trinity Large Thinking leads
GPT-4: 18.4 (#277), Trinity Large Thinking: 40.3 (#107)
| Benchmark | GPT-4 | Trinity Large Thinking |
|---|---|---|
| LMArena Expert | 1211 | 1359 |
| GPQA Diamond | 35.7% | — |
| Vectara Hallucination Rate | — | 6.9% |
| MMLU | 86.4% | — |
| TriviaQA | 84.8% | — |
Multilingual Trinity Large Thinking leads
GPT-4: 40.8 (#210), Trinity Large Thinking: 46.4 (#155)
| Benchmark | GPT-4 | Trinity Large Thinking |
|---|---|---|
| LMArena Non-English | 1246 | 1325 |
| LMArena Chinese | 1242 | 1372 |
| LMArena French | 1282 | 1374 |
| LMArena German | 1251 | 1356 |
| LMArena Japanese | 1209 | 1312 |
| LMArena Korean | 1184 | 1304 |
| LMArena Russian | 1251 | 1338 |
| LMArena Spanish | 1261 | 1356 |
Instruction Following Trinity Large Thinking leads
GPT-4: 65.5 (#217), Trinity Large Thinking: 70.7 (#157)
| Benchmark | GPT-4 | Trinity Large Thinking |
|---|---|---|
| LMArena Instruction Following | 1241 | 1334 |
Long Context Trinity Large Thinking leads
GPT-4: 37.7 (#207), Trinity Large Thinking: 41.2 (#139)
| Benchmark | GPT-4 | Trinity Large Thinking |
|---|---|---|
| LMArena Longer Query | 1244 | 1355 |
Writing & Preference Trinity Large Thinking leads
GPT-4: 35.0 (#263), Trinity Large Thinking: 54.0 (#152)
| Benchmark | GPT-4 | Trinity Large Thinking |
|---|---|---|
| LMArena Text | 1263 | 1340 |
| LMArena Creative Writing | 1244 | 1321 |
| LMArena Multi-Turn | 1257 | 1342 |
| EQ-Bench Creative Writing | 752 | — |
Frequently asked questions
Is GPT-4 better than Trinity Large Thinking?
Trinity Large Thinking is the stronger model overall, scoring 38.1 to 28.9 on the Noometry Index.
Which is cheaper, GPT-4 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; GPT-4 lists at $30 and $60.
Is GPT-4 or Trinity Large Thinking better for coding?
Trinity Large Thinking scores higher on coding benchmarks: 33.8 versus 31.2 in the Noometry coding category.
Which has the bigger context window?
Trinity Large Thinking does, with 262K tokens against 8K.
How many benchmarks do GPT-4 and Trinity Large Thinking share?
17 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and Trinity Large Thinking has 24.