Compare / Claude Sonnet 4.6 vs Trinity Large Thinking
Claude Sonnet 4.6 vs Trinity Large Thinking
Which is cheaper, which is more capable, and which does better at each kind of work, from 25 results on 3 sources that measured both models.
Claude Sonnet 4.6 or Trinity Large Thinking? The short answer
- Trinity Large Thinking is cheaper: $0.39 against $6.00 per million tokens, 15 times less.
- Claude Sonnet 4.6 scores higher on the Artificial Analysis Intelligence Index: 30.1 against 10.8.
- On our business benchmarks, Claude Sonnet 4.6 ranks higher for agents and tool use, reasoning and knowledge, writing and sticking to the facts.
At a glance
| Claude Sonnet 4.6 | Trinity Large Thinking | |
|---|---|---|
| Price per million tokens | $3.00 in · $15.00 out | $0.25 in · $0.80 out |
| Blended (3 in : 1 out) | $6.00 | $0.39 |
| Cheapest host, blended | $6.00 (Claude Platform on AWS) | – |
| Intelligence Index | 30.1 | 10.8 |
| Spring Prompt overall | – | – |
| Output speed, tokens a second | – | 344 |
| Context window | 1,000,000 tokens | 262,144 tokens |
| Weights | Closed | Open |
| Developer based in | the US | the US |
Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.
Which is better for what
Each model's rank among models available today, for each kind of work, from our own benchmarks and licensed sources. The better rank is in green.
| Use case | Claude Sonnet 4.6 | Trinity Large Thinking |
|---|---|---|
| Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls | 51 of 161 | 116 of 161 |
| Professional workReal tasks from banking, consulting and law, business documents and freelance projects | 31 of 51 | not measured |
| Reasoning and knowledgeHard questions across science, maths and general knowledge | 55 of 195 | 120 of 195 |
| WritingWhat people prefer in blind comparisons, and judged writing quality | 30 of 187 | 129 of 187 |
| Sticking to the factsSummarising without inventing things, and factual answers | 79 of 153 | 142 of 153 |
Every shared result
| Benchmark | Metric | Claude Sonnet 4.6 | Trinity Large Thinking |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,461 | 1,371 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,482 | 1,367 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,449 | 1,332 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,506 | 1,401 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,475 | 1,357 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,472 | 1,368 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,455 | 1,343 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 63.0 | 25.8 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 30.1 | 10.8 |
| Artificial Analysisreported | GPQA Diamond% of questions | 87.5% | 75.2% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 33.6% | 15.8% |
| Artificial Analysisreported | IFBench% of instructions | 56.6% | 56.3% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 80.0% | 38.0% |
| Artificial Analysisreported | SciCode% of problems | 50.1% | 40.6% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 71.2% | 20.6% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 3.0% | 0.5% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 53.0% | 22.7% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 34.4% | 5.8% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 79.5% | 90.1% |
| OpenHands Indexreported | Average% resolved | 44.5% | 32.1% |
| OpenHands Indexreported | Front end (SWE-Bench Multimodal)% resolved | 30.9% | 25.0% |
| OpenHands Indexreported | Greenfield (Commit0)% resolved | 50.0% | 12.5% |
| OpenHands Indexreported | Information gathering (GAIA)% resolved | 13.3% | 32.7% |
| OpenHands Indexreported | Issue resolution (SWE-Bench)% resolved | 74.4% | 56.8% |
| OpenHands Indexreported | Testing (SWT-Bench)% resolved | 54.0% | 33.3% |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.
Quick answers
Is Claude Sonnet 4.6 better than Trinity Large Thinking?
It depends on the task. Among models available today, on our benchmarks and licensed sources, Claude Sonnet 4.6 ranks higher for agents and tool use, reasoning and knowledge, writing and sticking to the facts.
Which is cheaper, Claude Sonnet 4.6 or Trinity Large Thinking?
Trinity Large Thinking costs $0.39 per million tokens (three input to one output), against $6.00 for Claude Sonnet 4.6.
Run Claude Sonnet 4.6 and Trinity Large Thinking on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.