Is Mistral Small 4 better than Trinity Large Thinking?

On accuracy they are statistically tied: Mistral Small 4 scored 32.5% and Trinity Large Thinking 28.8% on the same 80 questions from open datasets. Similar accuracy; Mistral Small 4 costs less per question.

Which is cheaper, Mistral Small 4 or Trinity Large Thinking?

Mistral Small 4 is cheaper: $0.0019 per question as billed, against $0.0025. List prices: Mistral Small 4 $0.15 in / $0.6 out, Trinity Large Thinking $0.25 in / $0.8 out per million tokens.

Which is faster, Mistral Small 4 or Trinity Large Thinking?

Trinity Large Thinking is faster at 329 output tokens per second, against 170. Time to first token: 0.5 s for Mistral Small 4, 0.6 s for Trinity Large Thinking.

Which has the larger context window, Mistral Small 4 or Trinity Large Thinking?

Both accept about the same amount of context: 262k and 262k tokens.

Head to head · tested 27 Oct 2026

Mistral Small 4 vs Trinity Large Thinking: accuracy, speed and price

Similar accuracy; Mistral Small 4 costs less per question.

MistralMistral Small 432.5% score · #62$0.0019 per question170 tok/s

Arcee AITrinity Large Thinking28.8% score · #65$0.0025 per question329 tok/s

Verdict

Built from the measurements below

Accuracy is too close to call: Mistral Small 4 scored 32.5% and Trinity Large Thinking 28.8% on the same 80 questions, a gap within what chance alone could produce.
Mistral Small 4 is 1.3× cheaper per question as billed ($0.0019 vs $0.0025), reasoning tokens included.
Trinity Large Thinking generates text 1.9× faster: 329 tokens per second against 170.
Mistral Small 4 starts answering sooner: 0.5 s to the first token, against 0.6 s for Trinity Large Thinking.

Which should you choose?

Choose Mistral Small 4 for the lower cost per correct answer ($0.0057 against $0.0087), a quicker first token and open weights you can run yourself. Choose Trinity Large Thinking for faster streaming. The widest subject gap is in medicine: Mistral Small 4 answered 5 of 11 correctly, Trinity Large Thinking 3 of 11. With so few questions per subject, treat this as a lead to test on your own prompts, not a result.

Accuracy gap +3.8 points (z = 0.51). A winner is called only when z is at least 1.96, the usual 95% threshold. Speed, latency and price need a 10% margin.

Head to head

Same questions, same prompt, same API

Measure	Mistral Small 4	Trinity Large Thinking	Edge
Verification scoreHigher is better	32.5%	28.8%	Too close to call
Cost per questionLower is better	$0.0019	$0.0025	Mistral Small 4
Output speedHigher is better	170	329	Trinity Large Thinking
Time to first tokenLower is better	0.5 s	0.6 s	Mistral Small 4
End-to-end response timeLower is better	4.5 s	2.7 s	Trinity Large Thinking
Tokens per questionLower is better	3,021	3,043	Similar
Context windowHigher is better	262k	262k	Similar

Side by side

Bars are scaled to the larger of the two values. The winner on each measure is marked in bold.

Every measure

Mistral Small 4 first, Trinity Large Thinking second

Verification scorehigher is better

Mistral Small 432.5%

Trinity Large Thinking28.8%

Cost per questionlower is better

Mistral Small 4$0.0019

Trinity Large Thinking$0.0025

Output speedhigher is better

Mistral Small 4170

Trinity Large Thinking329

Time to first tokenlower is better

Mistral Small 40.5 s

Trinity Large Thinking0.6 s

End-to-end response timelower is better

Mistral Small 44.5 s

Trinity Large Thinking2.7 s

Tokens per questionlower is better

Mistral Small 43,021

Trinity Large Thinking3,043

Context windowhigher is better

Mistral Small 4262k

Trinity Large Thinking262k

Score vs. cost per question

Both models in color, every other model in grey

Results by subject and dataset

Questions answered correctly. Small samples: a gap of one or two questions is noise.

Subject	Questions	Mistral Small 4	Trinity Large Thinking
Science	30	7/30	6/30
Engineering	15	0/15	0/15
Medicine	11	5/11	3/11
Economics	9	6/9	6/9
Law	6	2/6	2/6
Humanities & social science	9	6/9	6/9
SuperGPQA dataset	50	7/50	5/50
MMLU-Pro dataset	30	19/30	18/30

Specifications and pricing

From OpenRouter's catalogue, plus the leaderboard's test dates

Specification	Mistral Small 4	Trinity Large Thinking
Creator	Mistral	Arcee AI
Released	16 Mar 2026	1 Apr 2026
Weights	Open weights	—
Reasoning	Yes	Yes
Input	text, image	text
Output	text	text
Context window	262k tokens	262k tokens
Knowledge cutoff	—	—
List price in / out per 1M	$0.15 / $0.6	$0.25 / $0.8
Blended price per 1M (3:1)	$0.26	$0.39
Cheapest host	Mistral · $0.15 / $0.6	Arcee AI · $0.25 / $0.8
Hosts on OpenRouter	1	1
Last tested	26 Oct 2026	26 Oct 2026

Compare any two models →

Mistral Small 4 vs Trinity Large Thinking: frequently asked questions

Is Mistral Small 4 better than Trinity Large Thinking?: On accuracy they are statistically tied: Mistral Small 4 scored 32.5% and Trinity Large Thinking 28.8% on the same 80 questions from open datasets. Similar accuracy; Mistral Small 4 costs less per question.
Which is cheaper, Mistral Small 4 or Trinity Large Thinking?: Mistral Small 4 is cheaper: $0.0019 per question as billed, against $0.0025. List prices: Mistral Small 4 $0.15 in / $0.6 out, Trinity Large Thinking $0.25 in / $0.8 out per million tokens.
Which is faster, Mistral Small 4 or Trinity Large Thinking?: Trinity Large Thinking is faster at 329 output tokens per second, against 170. Time to first token: 0.5 s for Mistral Small 4, 0.6 s for Trinity Large Thinking.
Which has the larger context window, Mistral Small 4 or Trinity Large Thinking?: Both accept about the same amount of context: 262k and 262k tokens.