Claude Sonnet 4.6 vs Trinity Large Thinking
Wins 5 of 6 areas
Coding · Agents · Reasoning · Facts · Long documents
Wins 0 of 6 areas
—
Claude Sonnet 4.6 is the stronger all-rounder.Trinity Large Thinking is cheaper.
Scores updated · 22 tests both models report · How we compare
Where each one wins
Tests won in each of the six areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software40Claude Sonnet 4.64 of 4 tests
- AgentsCarrying out multi-step tasks on its own30Claude Sonnet 4.63 of 3 tests
- ReasoningHard problems that need careful thinking30Claude Sonnet 4.63 of 3 tests
- FactsGetting facts right instead of making them up21Claude Sonnet 4.62 of 3 tests
- Long documentsFinding answers in very long texts10Claude Sonnet 4.61 of 1 test
- Following instructionsDoing exactly what it is asked00Even0 each · 1 tie
Long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
Trinity Large Thinking costs 94% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Claude Sonnet 4.6 pulls ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+50.6points ahead
- Reasons across sets of long documentsAA-LCR+42points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+37.5points ahead
Where Trinity Large Thinking pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 22 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 4.6
- Terminal-Bench 2.1Claude Sonnet 4.6 by 50.671.220.6+50.6
- Terminal-Bench HardClaude Sonnet 4.6 by 30.35322.7+30.3
- LMArena · WebDevClaude Sonnet 4.6 by 283 rating points15221239+283 rating
- SciCodeClaude Sonnet 4.6 by 9.550.140.6+9.5
AgentsClaude Sonnet 4.6
- GDPValClaude Sonnet 4.6 by 36.736.70+36.7
- τ-Bench V3 · BankingClaude Sonnet 4.6 by 28.634.45.8+28.6
- Terminal-Bench 4.0Claude Sonnet 4.6 by 2.530.5+2.5
ReasoningClaude Sonnet 4.6
- Humanity's Last ExamClaude Sonnet 4.6 by 17.833.615.8+17.8
- GPQA DiamondClaude Sonnet 4.6 by 12.387.575.2+12.3
- CritPtClaude Sonnet 4.6 by 2.23.10.9+2.2
FactsClaude Sonnet 4.6
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.6 by 37.551.614.1+37.5
- AA-Omniscience · AccuracyClaude Sonnet 4.6 by 18.440.922.5+18.4
- Vectara HHEM hallucination ratelower is betterTrinity Large Thinking by 3.710.66.9+3.7
Other results7 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- AA-OmniscienceClaude Sonnet 4.6 by 56.312.2-44.1+56.3
- Artificial Analysis Coding IndexClaude Sonnet 4.6 by 37.26325.8+37.2
- AA IntelligenceClaude Sonnet 4.6 by 19.330.110.8+19.3
- τ²-Bench Telecom (AA run)Trinity Large Thinking by 14.475.790.1+14.4
- vectara_factual_consistencyTrinity Large Thinking by 3.789.493.1+3.7
- vectara_answer_ratetie99.999tie
- vectara_avg_summary_lengthtie114.7117.3tie
Questions people ask
Which is better, Claude Sonnet 4.6 or Trinity Large Thinking?
Claude Sonnet 4.6 wins five of the six areas where both have results: coding, agents, reasoning, facts and long documents. Trinity Large Thinking wins none, but costs 94% less. They are level on following instructions.
Which is better for coding?
Claude Sonnet 4.6. It wins 4 of the 4 coding tests both models report; Trinity Large Thinking wins none.
Which is cheaper?
Claude Sonnet 4.6 costs $3.00 per million input tokens and $15.00 per million output tokens; Trinity Large Thinking costs $0.25 and $0.80. That makes Trinity Large Thinking about 94% cheaper for the same work.
How do you compare the two?
We use the 22 benchmark tests both models have published scores on. The verdict counts the 15 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 7 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.