Is Mistral Medium 3.5 better than Qwen3.5 397B A17B?

On accuracy they are statistically tied: Mistral Medium 3.5 scored 28.8% and Qwen3.5 397B A17B 32.5% on the same 80 questions from open datasets. Similar accuracy; Qwen3.5 397B A17B costs less per question.

Which is cheaper, Mistral Medium 3.5 or Qwen3.5 397B A17B?

Qwen3.5 397B A17B is cheaper: $0.013 per question as billed, against $0.025. List prices: Mistral Medium 3.5 $1.5 in / $7.5 out, Qwen3.5 397B A17B $0.5 in / $3.6 out per million tokens.

Which is faster, Mistral Medium 3.5 or Qwen3.5 397B A17B?

Qwen3.5 397B A17B is faster at 240 output tokens per second, against 145. Time to first token: 0.5 s for Mistral Medium 3.5, 0.6 s for Qwen3.5 397B A17B.

Which has the larger context window, Mistral Medium 3.5 or Qwen3.5 397B A17B?

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

Head to head · tested 27 Oct 2026

Mistral Medium 3.5 vs Qwen3.5 397B A17B: accuracy, speed and price

Similar accuracy; Qwen3.5 397B A17B costs less per question.

MistralMistral Medium 3.528.8% score · #65$0.025 per question145 tok/s

QwenQwen3.5 397B A17B32.5% score · #62$0.013 per question240 tok/s

Verdict

Built from the measurements below

Accuracy is too close to call: Mistral Medium 3.5 scored 28.8% and Qwen3.5 397B A17B 32.5% on the same 80 questions, a gap within what chance alone could produce.
Qwen3.5 397B A17B is 1.9× cheaper per question as billed ($0.013 vs $0.025), reasoning tokens included.
Qwen3.5 397B A17B generates text 1.7× faster: 240 tokens per second against 145.
Mistral Medium 3.5 starts answering sooner: 0.5 s to the first token, against 0.6 s for Qwen3.5 397B A17B.

Which should you choose?

Choose Mistral Medium 3.5 for a quicker first token. Choose Qwen3.5 397B A17B for the lower cost per correct answer ($0.041 against $0.087), faster streaming, open weights you can run yourself and more hosts to choose from on OpenRouter (10 against 1). The widest subject gap is in science: Qwen3.5 397B A17B answered 10 of 30 correctly, Mistral Medium 3.5 5 of 30. 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 Medium 3.5	Qwen3.5 397B A17B	Edge
Verification scoreHigher is better	28.8%	32.5%	Too close to call
Cost per questionLower is better	$0.025	$0.013	Qwen3.5 397B A17B
Output speedHigher is better	145	240	Qwen3.5 397B A17B
Time to first tokenLower is better	0.5 s	0.6 s	Mistral Medium 3.5
End-to-end response timeLower is better	5.4 s	3.5 s	Qwen3.5 397B A17B
Tokens per questionLower is better	3,268	4,010	Mistral Medium 3.5
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 Medium 3.5 first, Qwen3.5 397B A17B second

Verification scorehigher is better

Mistral Medium 3.528.8%

Qwen3.5 397B A17B32.5%

Cost per questionlower is better

Mistral Medium 3.5$0.025

Qwen3.5 397B A17B$0.013

Output speedhigher is better

Mistral Medium 3.5145

Qwen3.5 397B A17B240

Time to first tokenlower is better

Mistral Medium 3.50.5 s

Qwen3.5 397B A17B0.6 s

End-to-end response timelower is better

Mistral Medium 3.55.4 s

Qwen3.5 397B A17B3.5 s

Tokens per questionlower is better

Mistral Medium 3.53,268

Qwen3.5 397B A17B4,010

Context windowhigher is better

Mistral Medium 3.5262k

Qwen3.5 397B A17B262k

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 Medium 3.5	Qwen3.5 397B A17B
Science	30	5/30	10/30
Engineering	15	0/15	0/15
Medicine	11	3/11	2/11
Economics	9	6/9	5/9
Law	6	2/6	1/6
Humanities & social science	9	7/9	8/9
SuperGPQA dataset	50	5/50	5/50
MMLU-Pro dataset	30	18/30	21/30

Specifications and pricing

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

Specification	Mistral Medium 3.5	Qwen3.5 397B A17B
Creator	Mistral	Qwen
Released	30 Apr 2026	16 Feb 2026
Weights	—	Open weights
Reasoning	Yes	Yes
Input	text, image, file	text, image, video
Output	text	text
Context window	262k tokens	262k tokens
Knowledge cutoff	—	—
List price in / out per 1M	$1.5 / $7.5	$0.5 / $3.6
Blended price per 1M (3:1)	$3.00	$1.27
Cheapest host	Mistral · $1.5 / $7.5	Alibaba · $0.39 / $2.34
Hosts on OpenRouter	1	10
Last tested	27 Oct 2026	27 Oct 2026

Compare any two models →

Mistral Medium 3.5 vs Qwen3.5 397B A17B: frequently asked questions

Is Mistral Medium 3.5 better than Qwen3.5 397B A17B?: On accuracy they are statistically tied: Mistral Medium 3.5 scored 28.8% and Qwen3.5 397B A17B 32.5% on the same 80 questions from open datasets. Similar accuracy; Qwen3.5 397B A17B costs less per question.
Which is cheaper, Mistral Medium 3.5 or Qwen3.5 397B A17B?: Qwen3.5 397B A17B is cheaper: $0.013 per question as billed, against $0.025. List prices: Mistral Medium 3.5 $1.5 in / $7.5 out, Qwen3.5 397B A17B $0.5 in / $3.6 out per million tokens.
Which is faster, Mistral Medium 3.5 or Qwen3.5 397B A17B?: Qwen3.5 397B A17B is faster at 240 output tokens per second, against 145. Time to first token: 0.5 s for Mistral Medium 3.5, 0.6 s for Qwen3.5 397B A17B.
Which has the larger context window, Mistral Medium 3.5 or Qwen3.5 397B A17B?: Both accept about the same amount of context: 262k and 262k tokens.