Model comparison
Gemma 3 27B vs Trinity Large Thinking
Trinity Large Thinking is the stronger model overall, scoring 38.6 to 30.8 on the Noometry Index. Gemma 3 27B costs 3.9× less per token, which makes it the better buy when Trinity Large Thinking's lead doesn't matter for your workload.
Last verified . 20 shared benchmarks.
Google
Arcee AISummary
- They share 20 benchmarks with published results for both. Gemma 3 27B scores higher in 2 categories and Trinity Large Thinking in 6 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Trinity Large Thinking leads 40.9 to 25.5.
- The biggest single-benchmark swing is SciCode: 21.2% for Gemma 3 27B and 36.1% for Trinity Large Thinking.
- Gemma 3 27B is cheaper at $0.08 / $0.16 per million input/output tokens, against $0.25 / $0.80 for Trinity Large Thinking.
- Trinity Large Thinking accepts more context: 262K tokens versus 131K.
Side by side
| Gemma 3 27B | Trinity Large Thinking | |
|---|---|---|
| Provider | Arcee AI | |
| Noometry Index | 30.8 | 38.6 |
| Released | 2025-03-11 | 2026-04-01 |
| Weights | Open | Open |
| Context window | 131K | 262K |
| Max output | 8K | 80K |
| Input $ / M tokens | $0.08 | $0.25 |
| Output $ / M tokens | $0.16 | $0.80 |
| Results tracked | 43 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Trinity Large Thinking leads
Gemma 3 27B: 22.5 (#334), Trinity Large Thinking: 34.1 (#244)
| Benchmark | Gemma 3 27B | Trinity Large Thinking |
|---|---|---|
| SciCode | 21.2% | 36.1% |
| LMArena Coding | 1322 | 1381 |
| Aider Polyglot | 4.9% | — |
| LMArena WebDev | — | 1238 |
| LiveBench Coding | 39.9% | — |
Agentic & Tool Use Not comparable
Gemma 3 27B: 25.1 (#110), Trinity Large Thinking: —
| Benchmark | Gemma 3 27B | Trinity Large Thinking |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | — |
Reasoning Too close to call
Gemma 3 27B: 16.7 (#301), Trinity Large Thinking: 16.9 (#298)
| Benchmark | Gemma 3 27B | Trinity Large Thinking |
|---|---|---|
| CritPt | 0% | 0.9% |
| LMArena Hard Prompts | 1340 | 1350 |
| Kagi LLM Benchmark | 40.4% | — |
| NYT Connections (extended) | — | 16.5% |
| Chess Puzzles | 0% | — |
| Thematic Generalization | — | 41.6% |
| LiveBench Reasoning | 43.8% | — |
| DTBench | 52.5% | — |
| LiveBench Data Analysis | 51.5% | — |
| LMCA | 12.3% | — |
| Surface Evolver Bench | — | 15.6% |
| Epoch Capabilities Index | 130.04 | — |
| LiveBench | 50% | — |
Math Trinity Large Thinking leads
Gemma 3 27B: 25.9 (#265), Trinity Large Thinking: 37.6 (#149)
| Benchmark | Gemma 3 27B | Trinity Large Thinking |
|---|---|---|
| LMArena Math | 1312 | 1366 |
| OTIS Mock AIME 2024-2025 | 22.5% | — |
| LiveBench Math | 55.4% | — |
| MATH Level 5 | 74% | — |
Knowledge Trinity Large Thinking leads
Gemma 3 27B: 25.5 (#261), Trinity Large Thinking: 40.9 (#113)
| Benchmark | Gemma 3 27B | Trinity Large Thinking |
|---|---|---|
| Vectara Hallucination Rate | 7.4% | 6.9% |
| LMArena Expert | 1304 | 1360 |
| GPQA Diamond | 47.7% | — |
| Confabulations | 40.3% | — |
Multimodal Not comparable
Gemma 3 27B: 32.6 (#100), Trinity Large Thinking: —
| Benchmark | Gemma 3 27B | Trinity Large Thinking |
|---|---|---|
| LMArena Vision | 1164 | — |
| GeoBench | 52% | — |
Multilingual Too close to call
Gemma 3 27B: 46.9 (#155), Trinity Large Thinking: 46.2 (#160)
| Benchmark | Gemma 3 27B | Trinity Large Thinking |
|---|---|---|
| LMArena Non-English | 1334 | 1325 |
| LMArena Chinese | 1346 | 1373 |
| LMArena French | 1368 | 1374 |
| LMArena German | 1362 | 1356 |
| LMArena Japanese | 1287 | 1311 |
| LMArena Korean | 1308 | 1306 |
| LMArena Russian | 1349 | 1337 |
| LMArena Spanish | 1349 | 1357 |
Instruction Following Too close to call
Gemma 3 27B: 70.6 (#160), Trinity Large Thinking: 70.5 (#162)
| Benchmark | Gemma 3 27B | Trinity Large Thinking |
|---|---|---|
| LMArena Instruction Following | 1321 | 1334 |
| LiveBench Instruction Following | 74.9% | — |
Long Context Trinity Large Thinking leads
Gemma 3 27B: 27.6 (#293), Trinity Large Thinking: 41.3 (#144)
| Benchmark | Gemma 3 27B | Trinity Large Thinking |
|---|---|---|
| LMArena Longer Query | 1333 | 1355 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference Trinity Large Thinking leads
Gemma 3 27B: 52.5 (#168), Trinity Large Thinking: 53.8 (#158)
| Benchmark | Gemma 3 27B | Trinity Large Thinking |
|---|---|---|
| LMArena Text | 1358 | 1340 |
| LMArena Creative Writing | 1346 | 1320 |
| LMArena Multi-Turn | 1345 | 1342 |
| Short-Story Creative Writing | 79.9% | — |
| EQ-Bench Creative Writing | 1266 | — |
| LiveBench Language | 34.6% | — |
Frequently asked questions
Is Gemma 3 27B better than Trinity Large Thinking?
Trinity Large Thinking is the stronger model overall, scoring 38.6 to 30.8 on the Noometry Index. Gemma 3 27B costs 3.9× less per token, which makes it the better buy when Trinity Large Thinking's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 27B or Trinity Large Thinking?
Gemma 3 27B is cheaper. It lists at $0.08 per million input tokens and $0.16 per million output tokens; Trinity Large Thinking lists at $0.25 and $0.80.
Is Gemma 3 27B or Trinity Large Thinking better for coding?
Trinity Large Thinking scores higher on coding benchmarks: 34.1 versus 22.5 in the Noometry coding category.
Which has the bigger context window?
Trinity Large Thinking does, with 262K tokens against 131K.
How many benchmarks do Gemma 3 27B and Trinity Large Thinking share?
20 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and Trinity Large Thinking has 24.