Model comparison
Step 5 Preview vs Trinity Large Thinking
Step 5 Preview is the stronger model overall, scoring 47.9 to 38.6 on the Noometry Index. Trinity Large Thinking costs 3.7× less per token, which makes it the better buy when Step 5 Preview's lead doesn't matter for your workload.
Last verified . 16 shared benchmarks.
StepFun
Arcee AISummary
- They share 16 benchmarks with published results for both. Step 5 Preview scores higher in 8 categories and Trinity Large Thinking in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Step 5 Preview leads 40.0 to 16.9.
- The biggest single-benchmark swing is SciCode: 58.9% for Step 5 Preview and 36.1% for Trinity Large Thinking.
- Trinity Large Thinking is cheaper at $0.25 / $0.80 per million input/output tokens, against $1 / $2.70 for Step 5 Preview.
- Step 5 Preview accepts more context: 1.02M tokens versus 262K.
- Trinity Large Thinking has downloadable open weights; the other is API-only.
Side by side
| Step 5 Preview | Trinity Large Thinking | |
|---|---|---|
| Provider | StepFun | Arcee AI |
| Noometry Index | 47.9 | 38.6 |
| Released | 2026-09-16 | 2026-04-01 |
| Weights | Proprietary | Open |
| Context window | 1.02M | 262K |
| Max output | 66K | 80K |
| Input $ / M tokens | $1 | $0.25 |
| Output $ / M tokens | $2.70 | $0.80 |
| Results tracked | 18 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 5 Preview leads
Step 5 Preview: 51.8 (#36), Trinity Large Thinking: 34.1 (#244)
| Benchmark | Step 5 Preview | Trinity Large Thinking |
|---|---|---|
| LMArena WebDev | 1564 | 1238 |
| SciCode | 58.9% | 36.1% |
| LMArena Coding | 1480 | 1381 |
Reasoning Step 5 Preview leads
Step 5 Preview: 40.0 (#57), Trinity Large Thinking: 16.9 (#298)
| Benchmark | Step 5 Preview | Trinity Large Thinking |
|---|---|---|
| CritPt | 20.9% | 0.9% |
| LMArena Hard Prompts | 1465 | 1350 |
| NYT Connections (extended) | — | 16.5% |
| Thematic Generalization | — | 41.6% |
| Surface Evolver Bench | — | 15.6% |
Math Step 5 Preview leads
Step 5 Preview: 46.1 (#72), Trinity Large Thinking: 37.6 (#149)
| Benchmark | Step 5 Preview | Trinity Large Thinking |
|---|---|---|
| LMArena Math | 1470 | 1366 |
| ProofBench | 42% | — |
Knowledge Too close to call
Step 5 Preview: 41.2 (#112), Trinity Large Thinking: 40.9 (#113)
| Benchmark | Step 5 Preview | Trinity Large Thinking |
|---|---|---|
| LMArena Expert | 1470 | 1360 |
| Vectara Hallucination Rate | — | 6.9% |
Multimodal Not comparable
Step 5 Preview: 41.0 (#41), Trinity Large Thinking: —
| Benchmark | Step 5 Preview | Trinity Large Thinking |
|---|---|---|
| LMArena Vision | 1267 | — |
Multilingual Step 5 Preview leads
Step 5 Preview: 53.8 (#54), Trinity Large Thinking: 46.2 (#160)
| Benchmark | Step 5 Preview | Trinity Large Thinking |
|---|---|---|
| LMArena Non-English | 1432 | 1325 |
| LMArena Chinese | 1519 | 1373 |
| LMArena Russian | 1436 | 1337 |
| LMArena Spanish | 1439 | 1357 |
| LMArena French | — | 1374 |
| LMArena German | — | 1356 |
| LMArena Japanese | — | 1311 |
| LMArena Korean | — | 1306 |
Instruction Following Step 5 Preview leads
Step 5 Preview: 76.0 (#50), Trinity Large Thinking: 70.5 (#162)
| Benchmark | Step 5 Preview | Trinity Large Thinking |
|---|---|---|
| LMArena Instruction Following | 1444 | 1334 |
Long Context Step 5 Preview leads
Step 5 Preview: 44.5 (#64), Trinity Large Thinking: 41.3 (#144)
| Benchmark | Step 5 Preview | Trinity Large Thinking |
|---|---|---|
| LMArena Longer Query | 1455 | 1355 |
Writing & Preference Step 5 Preview leads
Step 5 Preview: 62.8 (#71), Trinity Large Thinking: 53.8 (#158)
| Benchmark | Step 5 Preview | Trinity Large Thinking |
|---|---|---|
| LMArena Text | 1442 | 1340 |
| LMArena Creative Writing | 1410 | 1320 |
| LMArena Multi-Turn | 1447 | 1342 |
Frequently asked questions
Is Step 5 Preview better than Trinity Large Thinking?
Step 5 Preview is the stronger model overall, scoring 47.9 to 38.6 on the Noometry Index. Trinity Large Thinking costs 3.7× less per token, which makes it the better buy when Step 5 Preview's lead doesn't matter for your workload.
Which is cheaper, Step 5 Preview 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; Step 5 Preview lists at $1 and $2.70.
Is Step 5 Preview or Trinity Large Thinking better for coding?
Step 5 Preview scores higher on coding benchmarks: 51.8 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
Step 5 Preview does, with 1.02M tokens against 262K.
How many benchmarks do Step 5 Preview and Trinity Large Thinking share?
16 benchmarks have published results for both models. Step 5 Preview has 18 scored results on Noometry and Trinity Large Thinking has 24.