Model comparison
GPT-6 Astra vs Trinity Large Thinking
GPT-6 Astra is the stronger model overall, scoring 70.8 to 38.6 on the Noometry Index. Trinity Large Thinking costs 52× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
Arcee AISummary
- They share 22 benchmarks with published results for both. GPT-6 Astra scores higher in 8 categories and Trinity Large Thinking in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Astra leads 85.1 to 16.9.
- The biggest single-benchmark swing is NYT Connections (extended): 98.1% for GPT-6 Astra and 16.5% for Trinity Large Thinking.
- Trinity Large Thinking is cheaper at $0.25 / $0.80 per million input/output tokens, against $10 / $50 for GPT-6 Astra.
- GPT-6 Astra accepts more context: 1.05M tokens versus 262K.
- Trinity Large Thinking has downloadable open weights; the other is API-only.
Side by side
| GPT-6 Astra | Trinity Large Thinking | |
|---|---|---|
| Provider | OpenAI | Arcee AI |
| Noometry Index | 70.8 | 38.6 |
| Released | 2026-09-03 | 2026-04-01 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 262K |
| Max output | 128K | 80K |
| Input $ / M tokens | $10 | $0.25 |
| Output $ / M tokens | $50 | $0.80 |
| Results tracked | 56 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Astra leads
GPT-6 Astra: 73.7 (#2), Trinity Large Thinking: 34.1 (#244)
| Benchmark | GPT-6 Astra | Trinity Large Thinking |
|---|---|---|
| LMArena WebDev | 1786 | 1238 |
| SciCode | 56.5% | 36.1% |
| LMArena Coding | 1487 | 1381 |
| DeepSWE | 74.1% | — |
| FrontierCode | 53.3% | — |
| FrontierSWE | 65.5% | — |
| GSO | 79.4% | — |
| WeirdML | 93.6% | — |
| MirrorCode | 46.7% | — |
| ALE-Bench | 2,951 | — |
Agentic & Tool Use Not comparable
GPT-6 Astra: 52.9 (#3), Trinity Large Thinking: —
| Benchmark | GPT-6 Astra | Trinity Large Thinking |
|---|---|---|
| APEX-Agents | 64.7% | — |
| Remote Labor Index | 20.8% | — |
| BALROG | 68.3% | — |
| GDP.pdf | 34.2% | — |
| Vending-Bench 2 | 15,515 | — |
Reasoning GPT-6 Astra leads
GPT-6 Astra: 85.1 (#1), Trinity Large Thinking: 16.9 (#298)
| Benchmark | GPT-6 Astra | Trinity Large Thinking |
|---|---|---|
| NYT Connections (extended) | 98.1% | 16.5% |
| CritPt | 31.7% | 0.9% |
| LMArena Hard Prompts | 1462 | 1350 |
| ARC-AGI-2 | 95% | — |
| ARC-AGI-1 | 98.5% | — |
| Chess Puzzles | 72% | — |
| Thematic Generalization | — | 41.6% |
| EBR-Bench | 76.2% | — |
| Mystery Game Puzzles | 84% | — |
| DTBench | 97.3% | — |
| LMCA | 64.4% | — |
| Surface Evolver Bench | — | 15.6% |
| Bench to the Future 3 | 0.14 | — |
| Epoch Capabilities Index | 166.45 | — |
Math GPT-6 Astra leads
GPT-6 Astra: 93.5 (#2), Trinity Large Thinking: 37.6 (#149)
| Benchmark | GPT-6 Astra | Trinity Large Thinking |
|---|---|---|
| LMArena Math | 1465 | 1366 |
| FrontierMath (Tiers 1-3) | 93.7% | — |
| FrontierMath Tier 4 | 97.6% | — |
| OTIS Mock AIME 2024-2025 | 100% | — |
| ProofBench | 99% | — |
| FrontierMath Erdős | 2.9% | — |
Knowledge GPT-6 Astra leads
GPT-6 Astra: 75.3 (#1), Trinity Large Thinking: 40.9 (#113)
| Benchmark | GPT-6 Astra | Trinity Large Thinking |
|---|---|---|
| Vectara Hallucination Rate | 8.7% | 6.9% |
| LMArena Expert | 1483 | 1360 |
| GPQA Diamond | 95.8% | — |
| Humanity's Last Exam | 54.8% | — |
| SimpleQA Verified | 75.6% | — |
Multimodal Not comparable
GPT-6 Astra: 55.0 (#3), Trinity Large Thinking: —
| Benchmark | GPT-6 Astra | Trinity Large Thinking |
|---|---|---|
| LMArena Vision | 1281 | — |
| Blueprint-Bench 2 | 49.7% | — |
| Furniture Assembly | 80% | — |
| LMArena Document | 1468 | — |
Multilingual GPT-6 Astra leads
GPT-6 Astra: 53.7 (#61), Trinity Large Thinking: 46.2 (#160)
| Benchmark | GPT-6 Astra | Trinity Large Thinking |
|---|---|---|
| LMArena Non-English | 1430 | 1325 |
| LMArena Chinese | 1484 | 1373 |
| LMArena French | 1456 | 1374 |
| LMArena German | 1440 | 1356 |
| LMArena Japanese | 1379 | 1311 |
| LMArena Korean | 1426 | 1306 |
| LMArena Russian | 1436 | 1337 |
| LMArena Spanish | 1407 | 1357 |
Instruction Following GPT-6 Astra leads
GPT-6 Astra: 76.3 (#44), Trinity Large Thinking: 70.5 (#162)
| Benchmark | GPT-6 Astra | Trinity Large Thinking |
|---|---|---|
| LMArena Instruction Following | 1450 | 1334 |
Long Context GPT-6 Astra leads
GPT-6 Astra: 44.5 (#62), Trinity Large Thinking: 41.3 (#144)
| Benchmark | GPT-6 Astra | Trinity Large Thinking |
|---|---|---|
| LMArena Longer Query | 1456 | 1355 |
Writing & Preference GPT-6 Astra leads
GPT-6 Astra: 75.3 (#7), Trinity Large Thinking: 53.8 (#158)
| Benchmark | GPT-6 Astra | Trinity Large Thinking |
|---|---|---|
| LMArena Text | 1441 | 1340 |
| LMArena Creative Writing | 1418 | 1320 |
| LMArena Multi-Turn | 1448 | 1342 |
| EQ-Bench Creative Writing | 2173 | — |
Frequently asked questions
Is GPT-6 Astra better than Trinity Large Thinking?
GPT-6 Astra is the stronger model overall, scoring 70.8 to 38.6 on the Noometry Index. Trinity Large Thinking costs 52× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.
Which is cheaper, GPT-6 Astra 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-6 Astra lists at $10 and $50.
Is GPT-6 Astra or Trinity Large Thinking better for coding?
GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Astra does, with 1.05M tokens against 262K.
How many benchmarks do GPT-6 Astra and Trinity Large Thinking share?
22 benchmarks have published results for both models. GPT-6 Astra has 56 scored results on Noometry and Trinity Large Thinking has 24.