Model comparison
Solar Pro4 vs Trinity Large Thinking
Solar Pro4 is the stronger model overall, scoring 42.1 to 38.6 on the Noometry Index.
Last verified . 18 shared benchmarks.
Upstage
Arcee AISummary
- They share 18 benchmarks with published results for both. Solar Pro4 scores higher in 7 categories and Trinity Large Thinking in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Solar Pro4 leads 28.5 to 16.9.
- Trinity Large Thinking is cheaper at $0.25 / $0.80 per million input/output tokens, against $0.30 / $1.20 for Solar Pro4.
- Solar Pro4 accepts more context: 524K tokens versus 262K.
- Trinity Large Thinking has downloadable open weights; the other is API-only.
Side by side
| Solar Pro4 | Trinity Large Thinking | |
|---|---|---|
| Provider | Upstage | Arcee AI |
| Noometry Index | 42.1 | 38.6 |
| Released | 2026-08-06 | 2026-04-01 |
| Weights | Proprietary | Open |
| Context window | 524K | 262K |
| Max output | 131K | 80K |
| Input $ / M tokens | $0.30 | $0.25 |
| Output $ / M tokens | $1.20 | $0.80 |
| Results tracked | 18 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Solar Pro4 leads
Solar Pro4: 40.1 (#149), Trinity Large Thinking: 34.1 (#244)
| Benchmark | Solar Pro4 | Trinity Large Thinking |
|---|---|---|
| LMArena WebDev | 1371 | 1238 |
| LMArena Coding | 1437 | 1381 |
| SciCode | — | 36.1% |
Reasoning Solar Pro4 leads
Solar Pro4: 28.5 (#104), Trinity Large Thinking: 16.9 (#298)
| Benchmark | Solar Pro4 | Trinity Large Thinking |
|---|---|---|
| LMArena Hard Prompts | 1399 | 1350 |
| NYT Connections (extended) | — | 16.5% |
| CritPt | — | 0.9% |
| Thematic Generalization | — | 41.6% |
| Surface Evolver Bench | — | 15.6% |
Math Solar Pro4 leads
Solar Pro4: 38.8 (#128), Trinity Large Thinking: 37.6 (#149)
| Benchmark | Solar Pro4 | Trinity Large Thinking |
|---|---|---|
| LMArena Math | 1416 | 1366 |
Knowledge Trinity Large Thinking leads
Solar Pro4: 39.8 (#129), Trinity Large Thinking: 40.9 (#113)
| Benchmark | Solar Pro4 | Trinity Large Thinking |
|---|---|---|
| LMArena Expert | 1427 | 1360 |
| Vectara Hallucination Rate | — | 6.9% |
Multilingual Solar Pro4 leads
Solar Pro4: 48.7 (#139), Trinity Large Thinking: 46.2 (#160)
| Benchmark | Solar Pro4 | Trinity Large Thinking |
|---|---|---|
| LMArena Non-English | 1361 | 1325 |
| LMArena Chinese | 1415 | 1373 |
| LMArena French | 1397 | 1374 |
| LMArena German | 1364 | 1356 |
| LMArena Japanese | 1309 | 1311 |
| LMArena Korean | 1382 | 1306 |
| LMArena Russian | 1360 | 1337 |
| LMArena Spanish | 1401 | 1357 |
Instruction Following Solar Pro4 leads
Solar Pro4: 72.7 (#132), Trinity Large Thinking: 70.5 (#162)
| Benchmark | Solar Pro4 | Trinity Large Thinking |
|---|---|---|
| LMArena Instruction Following | 1377 | 1334 |
Long Context Too close to call
Solar Pro4: 42.1 (#130), Trinity Large Thinking: 41.3 (#144)
| Benchmark | Solar Pro4 | Trinity Large Thinking |
|---|---|---|
| LMArena Longer Query | 1381 | 1355 |
Writing & Preference Solar Pro4 leads
Solar Pro4: 56.5 (#138), Trinity Large Thinking: 53.8 (#158)
| Benchmark | Solar Pro4 | Trinity Large Thinking |
|---|---|---|
| LMArena Text | 1386 | 1340 |
| LMArena Creative Writing | 1316 | 1320 |
| LMArena Multi-Turn | 1385 | 1342 |
Frequently asked questions
Is Solar Pro4 better than Trinity Large Thinking?
Solar Pro4 is the stronger model overall, scoring 42.1 to 38.6 on the Noometry Index.
Which is cheaper, Solar Pro4 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; Solar Pro4 lists at $0.30 and $1.20.
Is Solar Pro4 or Trinity Large Thinking better for coding?
Solar Pro4 scores higher on coding benchmarks: 40.1 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
Solar Pro4 does, with 524K tokens against 262K.
How many benchmarks do Solar Pro4 and Trinity Large Thinking share?
18 benchmarks have published results for both models. Solar Pro4 has 18 scored results on Noometry and Trinity Large Thinking has 24.