Is Qwen3.5 397B A17B better than Qwen3.8 Max (0902)?

Qwen3.8 Max (0902) is more accurate: it answered 60 of 80 questions correctly (75.0%) against 26 of 80 (32.5%) for Qwen3.5 397B A17B. Qwen3.8 Max (0902) is more accurate and cheaper.

Which is cheaper, Qwen3.5 397B A17B or Qwen3.8 Max (0902)?

Qwen3.8 Max (0902) is cheaper: $0.011 per question as billed, against $0.013. List prices: Qwen3.5 397B A17B $0.5 in / $3.6 out, Qwen3.8 Max (0902) $2 in / $6 out per million tokens.

Which is faster, Qwen3.5 397B A17B or Qwen3.8 Max (0902)?

Qwen3.5 397B A17B is faster at 240 output tokens per second, against 46. Time to first token: 0.6 s for Qwen3.5 397B A17B, 1.2 s for Qwen3.8 Max (0902).

Which has the larger context window, Qwen3.5 397B A17B or Qwen3.8 Max (0902)?

Qwen3.8 Max (0902), with 1M tokens against 262k.

Head to head · tested 27 Oct 2026

Qwen3.5 397B A17B vs Qwen3.8 Max (0902): accuracy, speed and price

Qwen3.8 Max (0902) is more accurate and cheaper.

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

QwenQwen3.8 Max (0902)75.0% score · #15$0.011 per question46 tok/s

Verdict

Built from the measurements below

Qwen3.8 Max (0902) is more accurate: 75.0% against 32.5% on the same 80 questions, a 42.5-point gap that is statistically significant.
Qwen3.8 Max (0902) is 1.2× cheaper per question as billed ($0.011 vs $0.013), reasoning tokens included.
Qwen3.5 397B A17B generates text 5.2× faster: 240 tokens per second against 46.
Qwen3.5 397B A17B starts answering sooner: 0.6 s to the first token, against 1.2 s for Qwen3.8 Max (0902).
Qwen3.8 Max (0902) accepts more context: 1M tokens against 262k.

Which should you choose?

Choose Qwen3.5 397B A17B for faster streaming, a quicker first token, open weights you can run yourself and more hosts to choose from on OpenRouter (10 against 1). Choose Qwen3.8 Max (0902) for the most accurate answers of the two, the lower cost per correct answer ($0.015 against $0.041), shorter outputs (about 1,722 tokens per question against 4,010) and a 1M-token context window. The widest subject gap is in engineering: Qwen3.8 Max (0902) answered 9 of 15 correctly, Qwen3.5 397B A17B 0 of 15. With so few questions per subject, treat this as a lead to test on your own prompts, not a result.

Accuracy gap -42.5 points (z = -5.39). 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	Qwen3.5 397B A17B	Qwen3.8 Max (0902)	Edge
Verification scoreHigher is better	32.5%	75.0%	Qwen3.8 Max (0902)
Cost per questionLower is better	$0.013	$0.011	Qwen3.8 Max (0902)
Output speedHigher is better	240	46	Qwen3.5 397B A17B
Time to first tokenLower is better	0.6 s	1.2 s	Qwen3.5 397B A17B
End-to-end response timeLower is better	3.5 s	12.2 s	Qwen3.5 397B A17B
Tokens per questionLower is better	4,010	1,722	Qwen3.8 Max (0902)
Context windowHigher is better	262k	1M	Qwen3.8 Max (0902)

Side by side

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

Every measure

Qwen3.5 397B A17B first, Qwen3.8 Max (0902) second

Verification scorehigher is better

Qwen3.5 397B A17B32.5%

Qwen3.8 Max (0902)75.0%

Cost per questionlower is better

Qwen3.5 397B A17B$0.013

Qwen3.8 Max (0902)$0.011

Output speedhigher is better

Qwen3.5 397B A17B240

Qwen3.8 Max (0902)46

Time to first tokenlower is better

Qwen3.5 397B A17B0.6 s

Qwen3.8 Max (0902)1.2 s

End-to-end response timelower is better

Qwen3.5 397B A17B3.5 s

Qwen3.8 Max (0902)12.2 s

Tokens per questionlower is better

Qwen3.5 397B A17B4,010

Qwen3.8 Max (0902)1,722

Context windowhigher is better

Qwen3.5 397B A17B262k

Qwen3.8 Max (0902)1M

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	Qwen3.5 397B A17B	Qwen3.8 Max (0902)
Science	30	10/30	24/30
Engineering	15	0/15	9/15
Medicine	11	2/11	6/11
Economics	9	5/9	8/9
Law	6	1/6	4/6
Humanities & social science	9	8/9	9/9
SuperGPQA dataset	50	5/50	32/50
MMLU-Pro dataset	30	21/30	28/30

Specifications and pricing

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

Specification	Qwen3.5 397B A17B	Qwen3.8 Max (0902)
Creator	Qwen	Qwen
Released	16 Feb 2026	3 Sept 2026
Weights	Open weights	—
Reasoning	Yes	Yes
Input	text, image, video	text, image, video
Output	text	text
Context window	262k tokens	1M tokens
Knowledge cutoff	—	—
List price in / out per 1M	$0.5 / $3.6	$2 / $6
Blended price per 1M (3:1)	$1.27	$3.00
Cheapest host	Alibaba · $0.39 / $2.34	Alibaba · $2 / $6
Hosts on OpenRouter	10	1
Last tested	27 Oct 2026	25 Oct 2026

Compare any two models →

Qwen3.5 397B A17B vs Qwen3.8 Max (0902): frequently asked questions

Is Qwen3.5 397B A17B better than Qwen3.8 Max (0902)?: Qwen3.8 Max (0902) is more accurate: it answered 60 of 80 questions correctly (75.0%) against 26 of 80 (32.5%) for Qwen3.5 397B A17B. Qwen3.8 Max (0902) is more accurate and cheaper.
Which is cheaper, Qwen3.5 397B A17B or Qwen3.8 Max (0902)?: Qwen3.8 Max (0902) is cheaper: $0.011 per question as billed, against $0.013. List prices: Qwen3.5 397B A17B $0.5 in / $3.6 out, Qwen3.8 Max (0902) $2 in / $6 out per million tokens.
Which is faster, Qwen3.5 397B A17B or Qwen3.8 Max (0902)?: Qwen3.5 397B A17B is faster at 240 output tokens per second, against 46. Time to first token: 0.6 s for Qwen3.5 397B A17B, 1.2 s for Qwen3.8 Max (0902).
Which has the larger context window, Qwen3.5 397B A17B or Qwen3.8 Max (0902)?: Qwen3.8 Max (0902), with 1M tokens against 262k.