Compare / Qwen3.8-Max (0902) vs Qwen4-Max

Qwen3.8-Max (0902) vs Qwen4-Max

Which is cheaper, which is more capable, and which does better at each kind of work, from 57 results on 3 sources that measured both models.

AlibabaQwen3.8-Max (0902) Released 3 Aug 2026 · $2.00 in · $6.00 out
AlibabaQwen4-Max $1.50 in · $4.50 out
Qwen3.8-Max (0902) wins 1410 too close to callQwen4-Max wins 33

Of the 57 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.

Results as of 16 November 2026 (catalogue release 2026-11-16-49bdfa045dee).

Qwen3.8-Max (0902) or Qwen4-Max? The short answer

Pick Qwen4-Max: it scores higher (Spring Prompt overall 56 against 45) and costs less.

  • Spring Prompt overall: Qwen3.8-Max (0902) 45, Qwen4-Max 56 (out of 100; how the score works).
  • Qwen4-Max is cheaper to run: $0.007 against $0.026 per product listing on CatalogBench (3.7 times less).
  • On our own benchmarks, on the results both models have: Qwen4-Max does better for product listings, decks from an analysis and marketing planning.

Head to head

45
Spring Prompt overall
56
$3.00
Price per million tokens, blendedlower is better
$2.25
1,000,000
Context window, tokens
1,000,000

Every fact side by side

Qwen3.8-Max (0902)Qwen4-Max
Price per million tokens$2.00 in · $6.00 out$1.50 in · $4.50 out
Blended (3 in : 1 out)$3.00$2.25
Cheapest host, blended$3.00 (Alibaba)–
Cost per product listing (CatalogBench)$0.026$0.007
Intelligence Index45.4–
Spring Prompt overall4556
Output speed, tokens a second37–
Context window1,000,000 tokens1,000,000 tokens
Released3 Aug 2026–
WeightsClosedClosed
Developer based inChinaChina

Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.

Which is better for what

Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.

Product listingsTurning a sparse product feed and photos into listings that can go live 26 of 32 19 of 32
Decks from an analysisTurning a finished analysis into a deck you could present as it is 21 of 32 1 of 32
User surveysPlanning a user survey and reading its results without being misled not measured 5 of 23
Marketing planningPlanning a year of ad spend without overspending 27 of 32 15 of 32
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 10 of 159 not measured
Professional workReal tasks from banking, consulting and law, business documents and freelance projects 18 of 51 not measured
Reasoning and knowledgeHard questions across science, maths and general knowledge 21 of 187 not measured
WritingWhat people prefer in blind comparisons, and judged writing quality 17 of 169 not measured
Sticking to the factsSummarising without inventing things, and factual answers 36 of 139 not measured
Speed under pressureGood decisions against a real clock (fast chess) 36 of 36 26 of 35

From our research

Every shared result

Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.

MeasureQwen3.8-Max (0902)Qwen4-Max
CatalogBench measured by Spring Prompt · 3 results
Cost per productUS dollars, lower is better 25 · 13 of 33 measured $0.0256 $0.0069
Reliably publish-ready% of products, higher is better 22 · 19 of 32 measured 28.6% 39.3%
Reliably publish-ready · sales brief% of products, higher is better 27 · 21 of 32 measured 0.0% 8.9%
DeckBench measured by Spring Prompt · 3 results
Cost per deckUS dollars, lower is better 19 · 23 of 33 measured $0.15 $0.18
Deck ratingrating, higher is betterThe rating weighs each win by how strong the other deck was; the win rate counts wins against whichever decks a model happened to meet, so the two can disagree. 21 · 1 of 33 measured 1,035 1,585
Accurate decks% of tasks, higher is better 15 · 19 of 33 measured 50.0% 33.3%
ROASBench measured by Spring Prompt · 3 results
Overall scorescore out of 100, higher is better 28 · 15 of 33 measured 25.4 45.8
Audience scorescore out of 100, higher is better 31 · 23 of 33 measured 39.7 55.1
Business scorescore out of 100, higher is better 28 · 13 of 33 measured 15.6 48.5
Show all 57 results (48 more)
MeasureQwen3.8-Max (0902)Qwen4-Max
CatalogBench measured by Spring Prompt · 28 results
Channel compliance% of products, higher is better 1 · 1 of 33 measured 100.0% 100.0%
Channel rules broken% of products, lower is better 20 · 1 of 33 measured 0.6% 0.0%
Channel rules broken · sales brief% of products, lower is better 31 · 25 of 33 measured 3.0% 1.2%
Claims to check% of products, lower is better 1 · 27 of 31 measured 0.0% 1.8%
Claims to check · sales brief% of products, lower is better 22 · 23 of 31 measured 1.3% 3.0%
Conflicts caught% of conflicts, higher is better 1 · 1 of 33 measured 100.0% 100.0%
Content quality% of checks, higher is better 31 · 15 of 33 measured 94.2% 97.5%
Content quality · sales brief% of checks, higher is better 32 · 12 of 33 measured 93.6% 97.9%
Decision accuracy% of decisions, higher is better 29 · 17 of 33 measured 93.3% 96.8%
Failed outputs% of products, lower is better 32 · 1 of 33 measured 3.6% 0.0%
Failed outputs · sales brief% of products, lower is better 31 · 1 of 33 measured 3.6% 0.0%
Field accuracy% of missing fields, higher is better 32 · 21 of 33 measured 90.3% 92.9%
Invented values% of filled values, lower is better 26 · 27 of 33 measured 2.7% 2.9%
Missing UK information% of products, lower is better 1 · 1 of 33 measured 0.0% 0.0%
Missing UK information · sales brief% of products, lower is better 1 · 1 of 33 measured 0.0% 0.0%
Not findable% of products, lower is better 12 · 9 of 33 measured 3.6% 3.0%
Not findable · sales brief% of products, lower is better 5 · 20 of 33 measured 1.8% 4.2%
Publish-ready listings% of products, higher is better 22 · 20 of 33 measured 50.0% 56.6%
Publish-ready listings · sales brief% of products, higher is better 29 · 20 of 33 measured 7.7% 25.6%
Time per productseconds, lower is better 32 · 21 of 33 measured 78.4 20.9
Unsupported claimsclaims per product, lower is better 19 · 20 of 33 measured 0.69 0.70
Unsupported claims% of products, lower is better 21 · 20 of 33 measured 26.8% 22.0%
Unsupported claims · sales briefclaims per product, lower is better 32 · 23 of 33 measured 4.52 2.56
Unsupported claims · sales brief% of products, lower is better 31 · 22 of 33 measured 84.5% 70.2%
Wrong attributes% of products, lower is better 28 · 26 of 33 measured 21.4% 20.2%
Wrong attributes · sales brief% of products, lower is better 22 · 27 of 33 measured 19.6% 21.4%
Wrong category or variant% of products, lower is better 6 · 6 of 33 measured 1.8% 1.8%
Wrong category or variant · sales brief% of products, lower is better 18 · 18 of 33 measured 2.4% 2.4%
DeckBench measured by Spring Prompt · 14 results
Caveat dropped% of tasks, lower is better 15 · 15 of 33 measured 16.7% 16.7%
Clean layout% of tasks, higher is better 8 · 24 of 33 measured 66.7% 16.7%
Design quality% of the maximum, higher is better 16 · 4 of 33 measured 67.0% 83.4%
Draft figure quoted% of tasks, lower is better 19 · 13 of 33 measured 33.3% 16.7%
Findings missing% of tasks, lower is better 1 · 1 of 33 measured 0.0% 0.0%
Head-to-head win rate% of comparisons, higher is betterThe rating weighs each win by how strong the other deck was; the win rate counts wins against whichever decks a model happened to meet, so the two can disagree. 16 · 3 of 33 measured 60.8% 86.3%
Layout defects% of tasks, lower is better 8 · 24 of 33 measured 33.3% 83.3%
Misleading metric used% of tasks, lower is better 1 · 1 of 33 measured 0.0% 0.0%
Presentable decks% of tasks, higher is better 7 · 15 of 33 measured 33.3% 0.0%
Recommendation late or wrong% of tasks, lower is better 1 · 1 of 33 measured 0.0% 0.0%
Slides needing work% of tasks, lower is better 17 · 10 of 33 measured 50.0% 33.3%
Time per deckseconds, lower is better 32 · 31 of 33 measured 543 489
Unsupported claims% of tasks, lower is better 15 · 24 of 33 measured 16.7% 33.3%
Unsupported numbers% of tasks, lower is better 1 · 26 of 33 measured 0.0% 16.7%
ROASBench measured by Spring Prompt · 6 results
Consistency scorescore out of 100, higher is better 25 · 22 of 33 measured 19.3 25.5
Contribution profitsimulated US dollars, higher is better 28 · 7 of 33 measured $99,775 $775,520
Cost of a runUS dollars, lower is better 27 · 26 of 33 measured $0.73 $0.69
Months over budgetmonths of 12, lower is better 31 · 1 of 33 measured 2 0
Planning scorescore out of 100, higher is better 17 · 31 of 33 measured 54.8 54.0
Return on ad spendprofit per $1 spent, higher is better 28 · 7 of 33 measured $0.09 $0.68

Quick answers

Is Qwen3.8-Max (0902) better than Qwen4-Max?

Pick Qwen4-Max: it scores higher (Spring Prompt overall 56 against 45) and costs less.

Which is cheaper, Qwen3.8-Max (0902) or Qwen4-Max?

Qwen4-Max is cheaper to run: $0.007 against $0.026 per product listing on CatalogBench (3.7 times less).

Which is better for product listings, Qwen3.8-Max (0902) or Qwen4-Max?

Qwen4-Max ranks 19 of 32 models on sale for turning a sparse product feed and photos into listings that can go live, against 26 for Qwen3.8-Max (0902).

Which is better for decks from an analysis, Qwen3.8-Max (0902) or Qwen4-Max?

Qwen4-Max ranks 1 of 32 models on sale for turning a finished analysis into a deck you could present as it is, against 21 for Qwen3.8-Max (0902).

Which is better for marketing planning, Qwen3.8-Max (0902) or Qwen4-Max?

Qwen4-Max ranks 15 of 32 models on sale for planning a year of ad spend without overspending, against 27 for Qwen3.8-Max (0902).

Which is better for speed under pressure, Qwen3.8-Max (0902) or Qwen4-Max?

Qwen4-Max ranks 26 of 36 models on sale for good decisions against a real clock (fast chess), against 36 for Qwen3.8-Max (0902).

Run Qwen3.8-Max (0902) and Qwen4-Max on your own prompt.

Benchmarks aren't your task. Test both side by side on your prompt, with every answer checked against your rules and the cost of each one.