Model comparison
o4-mini vs Trinity Large Thinking
o4-mini is the stronger model overall, scoring 41.6 to 38.6 on the Noometry Index. Trinity Large Thinking costs 5.0× less per token, which makes it the better buy when o4-mini's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
Arcee AISummary
- They share 19 benchmarks with published results for both. o4-mini scores higher in 8 categories and Trinity Large Thinking in 0 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where o4-mini leads 24.6 to 16.9.
- The biggest single-benchmark swing is Vectara Hallucination Rate: 18.6% for o4-mini and 6.9% for Trinity Large Thinking.
- Trinity Large Thinking is cheaper at $0.25 / $0.80 per million input/output tokens, against $1.10 / $4.40 for o4-mini.
- Trinity Large Thinking accepts more context: 262K tokens versus 200K.
- Trinity Large Thinking has downloadable open weights; the other is API-only.
Side by side
| o4-mini | Trinity Large Thinking | |
|---|---|---|
| Provider | OpenAI | Arcee AI |
| Noometry Index | 41.6 | 38.6 |
| Released | 2025-04-16 | 2026-04-01 |
| Weights | Proprietary | Open |
| Context window | 200K | 262K |
| Max output | 100K | 80K |
| Input $ / M tokens | $1.10 | $0.25 |
| Output $ / M tokens | $4.40 | $0.80 |
| Results tracked | 60 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o4-mini leads
o4-mini: 40.9 (#127), Trinity Large Thinking: 34.1 (#244)
| Benchmark | o4-mini | Trinity Large Thinking |
|---|---|---|
| LMArena Coding | 1368 | 1381 |
| SWE-bench Verified (bash only) | 45% | — |
| Aider Polyglot | 72% | — |
| LMArena WebDev | — | 1238 |
| SciCode | — | 36.1% |
| GSO | 3.6% | — |
| WeirdML | 52.6% | — |
| CadEval | 62% | — |
| ALE-Bench | 826.17 | — |
| AlgoTune | 1.72 | — |
Agentic & Tool Use Not comparable
o4-mini: 32.6 (#61), Trinity Large Thinking: —
| Benchmark | o4-mini | Trinity Large Thinking |
|---|---|---|
| Berkeley Function Calling Leaderboard | 53.2% | — |
| GDPval | 25.3% | — |
| METR Time Horizons | 63.9% | — |
Reasoning o4-mini leads
o4-mini: 24.6 (#162), Trinity Large Thinking: 16.9 (#298)
| Benchmark | o4-mini | Trinity Large Thinking |
|---|---|---|
| CritPt | 0.6% | 0.9% |
| LMArena Hard Prompts | 1351 | 1350 |
| ARC-AGI-2 | 6.1% | — |
| SimpleBench | 38.7% | — |
| Kagi LLM Benchmark | 67.6% | — |
| NYT Connections (extended) | — | 16.5% |
| ARC-AGI-1 | 58.7% | — |
| Chess Puzzles | 26% | — |
| EnigmaEval | 9.2% | — |
| Thematic Generalization | — | 41.6% |
| Mystery Game Puzzles | 5% | — |
| DTBench | 77.6% | — |
| LMCA | 26.5% | — |
| Surface Evolver Bench | — | 15.6% |
| Epoch Capabilities Index | 145.64 | — |
| ForecastBench | 61.8 | — |
Math o4-mini leads
o4-mini: 40.8 (#89), Trinity Large Thinking: 37.6 (#149)
| Benchmark | o4-mini | Trinity Large Thinking |
|---|---|---|
| LMArena Math | 1389 | 1366 |
| FrontierMath (Tiers 1-3) | 36.1% | — |
| FrontierMath Tier 4 | 4.9% | — |
| OTIS Mock AIME 2024-2025 | 81.7% | — |
| Omni-MATH | 72% | — |
| MATH Level 5 | 97.8% | — |
| FrontierMath (Feb 2025 set) | 24.8% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge o4-mini leads
o4-mini: 43.6 (#91), Trinity Large Thinking: 40.9 (#113)
| Benchmark | o4-mini | Trinity Large Thinking |
|---|---|---|
| Vectara Hallucination Rate | 18.6% | 6.9% |
| LMArena Expert | 1343 | 1360 |
| GPQA Diamond | 79.6% | — |
| Humanity's Last Exam | 18.1% | — |
| SimpleQA Verified | 19.6% | — |
| MMLU-Pro | 82% | — |
| Confabulations | 15.8% | — |
| GPQA (HELM) | 73.5% | — |
Multimodal Not comparable
o4-mini: 40.2 (#49), Trinity Large Thinking: —
| Benchmark | o4-mini | Trinity Large Thinking |
|---|---|---|
| LMArena Vision | 1194 | — |
| GeoBench | 64% | — |
| VPCT | 57.5% | — |
Multilingual Too close to call
o4-mini: 47.0 (#154), Trinity Large Thinking: 46.2 (#160)
| Benchmark | o4-mini | Trinity Large Thinking |
|---|---|---|
| LMArena Non-English | 1337 | 1325 |
| LMArena Chinese | 1354 | 1373 |
| LMArena French | 1364 | 1374 |
| LMArena German | 1336 | 1356 |
| LMArena Japanese | 1308 | 1311 |
| LMArena Korean | 1312 | 1306 |
| LMArena Russian | 1334 | 1337 |
| LMArena Spanish | 1347 | 1357 |
Instruction Following o4-mini leads
o4-mini: 75.2 (#68), Trinity Large Thinking: 70.5 (#162)
| Benchmark | o4-mini | Trinity Large Thinking |
|---|---|---|
| LMArena Instruction Following | 1321 | 1334 |
| IFEval | 92.8% | — |
Long Context o4-mini leads
o4-mini: 45.5 (#33), Trinity Large Thinking: 41.3 (#144)
| Benchmark | o4-mini | Trinity Large Thinking |
|---|---|---|
| LMArena Longer Query | 1315 | 1355 |
| Fiction.LiveBench | 77.8% | — |
Writing & Preference Too close to call
o4-mini: 54.0 (#152), Trinity Large Thinking: 53.8 (#158)
| Benchmark | o4-mini | Trinity Large Thinking |
|---|---|---|
| LMArena Text | 1353 | 1340 |
| LMArena Creative Writing | 1294 | 1320 |
| LMArena Multi-Turn | 1350 | 1342 |
| Short-Story Creative Writing | 75% | — |
| WildBench | 85.4% | — |
Frequently asked questions
Is o4-mini better than Trinity Large Thinking?
o4-mini is the stronger model overall, scoring 41.6 to 38.6 on the Noometry Index. Trinity Large Thinking costs 5.0× less per token, which makes it the better buy when o4-mini's lead doesn't matter for your workload.
Which is cheaper, o4-mini 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; o4-mini lists at $1.10 and $4.40.
Is o4-mini or Trinity Large Thinking better for coding?
o4-mini scores higher on coding benchmarks: 40.9 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
Trinity Large Thinking does, with 262K tokens against 200K.
How many benchmarks do o4-mini and Trinity Large Thinking share?
19 benchmarks have published results for both models. o4-mini has 60 scored results on Noometry and Trinity Large Thinking has 24.