Is Qwen3.8 Max (0902) better than Qwen3 Coder Next?

On accuracy they are statistically tied: Qwen3.8 Max (0902) scored 75.0% and Qwen3 Coder Next 62.5% on the same 80 questions from open datasets. Similar accuracy; Qwen3 Coder Next costs less per question.

Which is cheaper, Qwen3.8 Max (0902) or Qwen3 Coder Next?

Qwen3 Coder Next is cheaper: $0.0017 per question as billed, against $0.011. List prices: Qwen3.8 Max (0902) $2 in / $6 out, Qwen3 Coder Next $0.12 in / $0.8 out per million tokens.

Which is faster, Qwen3.8 Max (0902) or Qwen3 Coder Next?

Qwen3 Coder Next is faster at 106 output tokens per second, against 46. Time to first token: 1.2 s for Qwen3.8 Max (0902), 0.4 s for Qwen3 Coder Next.

Which has the larger context window, Qwen3.8 Max (0902) or Qwen3 Coder Next?

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

Head to head · tested 27 Oct 2026

Qwen3.8 Max (0902) vs Qwen3 Coder Next: accuracy, speed and price

Similar accuracy; Qwen3 Coder Next costs less per question.

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

QwenQwen3 Coder Next62.5% score · #36$0.0017 per question106 tok/s

Verdict

Built from the measurements below

Accuracy is too close to call: Qwen3.8 Max (0902) scored 75.0% and Qwen3 Coder Next 62.5% on the same 80 questions, a gap within what chance alone could produce.
Qwen3 Coder Next is 6.6× cheaper per question as billed ($0.0017 vs $0.011), reasoning tokens included.
Qwen3 Coder Next generates text 2.3× faster: 106 tokens per second against 46.
Qwen3 Coder Next starts answering sooner: 0.4 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 Qwen3.8 Max (0902) for a 1M-token context window. Choose Qwen3 Coder Next for the lower cost per correct answer ($0.0027 against $0.015), faster streaming, a quicker first token and open weights you can run yourself. The widest subject gap is in engineering: Qwen3.8 Max (0902) answered 9 of 15 correctly, Qwen3 Coder Next 6 of 15. With so few questions per subject, treat this as a lead to test on your own prompts, not a result.

Accuracy gap +12.5 points (z = 1.71). 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	Qwen3.8 Max (0902)	Qwen3 Coder Next	Edge
Verification scoreHigher is better	75.0%	62.5%	Too close to call
Cost per questionLower is better	$0.011	$0.0017	Qwen3 Coder Next
Output speedHigher is better	46	106	Qwen3 Coder Next
Time to first tokenLower is better	1.2 s	0.4 s	Qwen3 Coder Next
End-to-end response timeLower is better	12.2 s	4.2 s	Qwen3 Coder Next
Tokens per questionLower is better	1,722	2,059	Qwen3.8 Max (0902)
Context windowHigher is better	1M	262k	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

Qwen3.8 Max (0902) first, Qwen3 Coder Next second

Verification scorehigher is better

Qwen3.8 Max (0902)75.0%

Qwen3 Coder Next62.5%

Cost per questionlower is better

Qwen3.8 Max (0902)$0.011

Qwen3 Coder Next$0.0017

Output speedhigher is better

Qwen3.8 Max (0902)46

Qwen3 Coder Next106

Time to first tokenlower is better

Qwen3.8 Max (0902)1.2 s

Qwen3 Coder Next0.4 s

End-to-end response timelower is better

Qwen3.8 Max (0902)12.2 s

Qwen3 Coder Next4.2 s

Tokens per questionlower is better

Qwen3.8 Max (0902)1,722

Qwen3 Coder Next2,059

Context windowhigher is better

Qwen3.8 Max (0902)1M

Qwen3 Coder Next262k

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	Qwen3.8 Max (0902)	Qwen3 Coder Next
Science	30	24/30	19/30
Engineering	15	9/15	6/15
Medicine	11	6/11	5/11
Economics	9	8/9	7/9
Law	6	4/6	4/6
Humanities & social science	9	9/9	9/9
SuperGPQA dataset	50	32/50	23/50
MMLU-Pro dataset	30	28/30	27/30

Specifications and pricing

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

Specification	Qwen3.8 Max (0902)	Qwen3 Coder Next
Creator	Qwen	Qwen
Released	3 Sept 2026	4 Feb 2026
Weights	—	Open weights
Reasoning	Yes	No
Input	text, image, video	text
Output	text	text
Context window	1M tokens	262k tokens
Knowledge cutoff	—	—
List price in / out per 1M	$2 / $6	$0.12 / $0.8
Blended price per 1M (3:1)	$3.00	$0.29
Cheapest host	Alibaba · $2 / $6	Parasail · $0.12 / $0.8
Hosts on OpenRouter	1	1
Last tested	25 Oct 2026	26 Oct 2026

Compare any two models →

Qwen3.8 Max (0902) vs Qwen3 Coder Next: frequently asked questions

Is Qwen3.8 Max (0902) better than Qwen3 Coder Next?: On accuracy they are statistically tied: Qwen3.8 Max (0902) scored 75.0% and Qwen3 Coder Next 62.5% on the same 80 questions from open datasets. Similar accuracy; Qwen3 Coder Next costs less per question.
Which is cheaper, Qwen3.8 Max (0902) or Qwen3 Coder Next?: Qwen3 Coder Next is cheaper: $0.0017 per question as billed, against $0.011. List prices: Qwen3.8 Max (0902) $2 in / $6 out, Qwen3 Coder Next $0.12 in / $0.8 out per million tokens.
Which is faster, Qwen3.8 Max (0902) or Qwen3 Coder Next?: Qwen3 Coder Next is faster at 106 output tokens per second, against 46. Time to first token: 1.2 s for Qwen3.8 Max (0902), 0.4 s for Qwen3 Coder Next.
Which has the larger context window, Qwen3.8 Max (0902) or Qwen3 Coder Next?: Qwen3.8 Max (0902), with 1M tokens against 262k.