Compare / Qwen3.8-Max (0902) vs Claude Sonnet 5.5
Qwen3.8-Max (0902) vs Claude Sonnet 5.5
Which is cheaper, which is more capable, and which does better at each kind of work, from 74 results on 6 sources that measured both models.
AlibabaQwen3.8-Max (0902)
AnthropicClaude Sonnet 5.5
Qwen3.8-Max (0902) wins 1114 too close to callClaude Sonnet 5.5 wins 49
Of the 74 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.
Results as of 9 October 2026 (catalogue release 2026-10-09-dd620e778ac5). See also Alibaba vs Anthropic, every model from both labs.
Qwen3.8-Max (0902) or Claude Sonnet 5.5? The short answer
Pick Claude Sonnet 5.5: it scores higher (Spring Prompt overall 64 against 48) and costs less.
- Spring Prompt overall: Qwen3.8-Max (0902) 48, Claude Sonnet 5.5 64 (out of 100; how the score works).
- Claude Sonnet 5.5 is cheaper to run: $0.017 against $0.026 per product listing on CatalogBench (33% less).
- Claude Sonnet 5.5 scores higher on the Artificial Analysis Intelligence Index: 56.0 against 45.4.
- Claude Sonnet 5.5 writes faster: 134 tokens a second against 37.
- Claude Sonnet 5.5 is newer: released 28 Sep 2026, against 3 Aug 2026 for Qwen3.8-Max (0902).
- On our own benchmarks, on the results both models have: Claude Sonnet 5.5 does better for product listings, decks from an analysis and marketing planning.
- On third-party benchmarks, on the results both models have: Qwen3.8-Max (0902) does better for sticking to the facts; Claude Sonnet 5.5 for agents and tool use and reasoning and knowledge.
Head to head
48
Spring Prompt overall
64
45.4
Intelligence Index
56.0
$3.00
Price per million tokens, blendedlower is better
$4.00
37
Output speed, tokens a second
134
1,000,000
Context window, tokens
1,000,000
Every fact side by side
| Qwen3.8-Max (0902) | Claude Sonnet 5.5 | |
|---|---|---|
| Price per million tokens | $2.00 in · $6.00 out | $2.00 in · $10.00 out |
| Blended (3 in : 1 out) | $3.00 | $4.00 |
| Cheapest host, blended | – | $4.00 (Amazon Bedrock) |
| Cost per product listing (CatalogBench) | $0.026 | $0.017 |
| Cached input, per million | – | $0.20 cached input |
| Intelligence Index | 45.4 | 56.0 |
| Spring Prompt overall | 48 | 64 |
| Output speed, tokens a second | 37 | 134 |
| Context window | 1,000,000 tokens | 1,000,000 tokens |
| Released | 3 Aug 2026 | 28 Sep 2026 (newer) |
| Weights | Closed | Closed |
| Developer based in | China | the US |
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
17 of 20
10 of 20
Decks from an analysisTurning a finished analysis into a deck you could present as it is
11 of 19
4 of 19
Marketing planningPlanning a year of ad spend without overspending
14 of 19
4 of 19
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls
7 of 164
1 of 164
Professional workReal tasks from banking, consulting and law, business documents and freelance projects
18 of 51
1 of 51
Reasoning and knowledgeHard questions across science, maths and general knowledge
17 of 198
2 of 198
WritingWhat people prefer in blind comparisons, and judged writing quality
16 of 188
32 of 188
Sticking to the factsSummarising without inventing things, and factual answers
38 of 152
42 of 152
Speed under pressureGood decisions against a real clock (fast chess)
24 of 24
23 of 23
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.
| Measure | Qwen3.8-Max (0902) | Claude Sonnet 5.5 |
|---|---|---|
| CatalogBench measured by Spring Prompt · 3 results | ||
| Cost per productUS dollars, lower is better 15 · 13 of 20 measured | $0.0256 | $0.0172 |
| Reliably publish-ready% of products, higher is better 13 · 12 of 20 measured | 28.6% | 32.1% |
| Reliably publish-ready · sales brief% of products, higher is better 18 · 9 of 20 measured | 0.0% | 10.7% |
| DeckBench measured by Spring Prompt · 3 results | ||
| Cost per deckUS dollars, lower is better 11 · 12 of 19 measured | $0.15 | $0.15 |
| 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. 11 · 4 of 19 measured | 1,100 | 1,352 |
| Accurate decks% of tasks, higher is better 8 · 5 of 19 measured | 50.0% | 66.7% |
| ROASBench measured by Spring Prompt · 3 results | ||
| Overall scorescore out of 100, higher is better 14 · 4 of 19 measured | 25.4 | 52.0 |
| Audience scorescore out of 100, higher is better 17 · 9 of 19 measured | 39.7 | 61.8 |
| Business scorescore out of 100, higher is better 14 · 2 of 19 measured | 15.6 | 57.2 |
| Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results | ||
| Overall · TextArena rating, higher is better 20 · 39 of 387 measured | 1,482 | 1,471xhigh reasoning |
| Overall · Text factualityArena rating, higher is better 22 · 24 of 163 measured | 1,472 | 1,470xhigh reasoning |
| Business, management and finance · TextArena rating, higher is better 16 · 46 of 380 measured | 1,484 | 1,464xhigh reasoning |
| Artificial Analysis reported by Artificial Analysis · 3 results | ||
| Artificial Analysis Intelligence Indexindex score, higher is better 14 · 2 of 243 measured | 45.4 | 56.0max reasoning |
| Humanity's Last Exam% of questions, higher is better 24 · 5 of 242 measured | 43.1% | 55.0%max reasoning |
| Terminal-Bench 4.0% of tasks, higher is better 12 · 1 of 112 measured | 38.9% | 63.6%max reasoning |
| SimpleQA Verified (Epoch AI) reported by SimpleQA Verified (Epoch AI) · 1 result | ||
| Correct answers% of questions, higher is better 30 · 33 of 76 measured | 47.3%xhigh reasoning | 46.5%max reasoning |
Show all 74 results (58 more)
| Measure | Qwen3.8-Max (0902) | Claude Sonnet 5.5 |
|---|---|---|
| CatalogBench measured by Spring Prompt · 27 results | ||
| Channel compliance% of products, higher is better 1 · 14 of 20 measured | 100.0% | 99.4% |
| Channel rules broken% of products, lower is better 11 · 16 of 20 measured | 0.6% | 1.2% |
| Channel rules broken · sales brief% of products, lower is better 19 · 9 of 20 measured | 3.0% | 0.6% |
| Claims to check% of products, lower is better 1 · 1 of 20 measured | 0.0% | 0.0% |
| Claims to check · sales brief% of products, lower is better 13 · 11 of 20 measured | 1.3% | 0.6% |
| Conflicts caught% of conflicts, higher is better 1 · 1 of 20 measured | 100.0% | 100.0% |
| Content quality% of checks, higher is better 18 · 13 of 20 measured | 94.2% | 96.5% |
| Content quality · sales brief% of checks, higher is better 19 · 9 of 20 measured | 93.6% | 97.3% |
| Decision accuracy% of decisions, higher is better 16 · 13 of 20 measured | 93.3% | 95.7% |
| Failed outputs% of products, lower is better 19 · 1 of 20 measured | 3.6% | 0.0% |
| Failed outputs · sales brief% of products, lower is better 19 · 1 of 20 measured | 3.6% | 0.0% |
| Field accuracy% of missing fields, higher is better 19 · 17 of 20 measured | 90.3% | 91.5% |
| Invented values% of filled values, lower is better 15 · 7 of 20 measured | 2.7% | 0.8% |
| Missing UK information% of products, lower is better 1 · 1 of 20 measured | 0.0% | 0.0% |
| Missing UK information · sales brief% of products, lower is better 1 · 1 of 20 measured | 0.0% | 0.0% |
| Not findable% of products, lower is better 4 · 1 of 20 measured | 3.6% | 1.2% |
| Not findable · sales brief% of products, lower is better 2 · 1 of 20 measured | 1.8% | 0.6% |
| Publish-ready listings% of products, higher is better 12 · 14 of 20 measured | 50.0% | 46.4% |
| Publish-ready listings · sales brief% of products, higher is better 18 · 10 of 20 measured | 7.7% | 27.4% |
| Unsupported claims% of products, lower is better 10 · 16 of 20 measured | 26.8% | 39.3% |
| Unsupported claimsclaims per product, lower is better 10 · 14 of 20 measured | 0.69 | 1.08 |
| Unsupported claims · sales briefclaims per product, lower is better 19 · 10 of 20 measured | 4.52 | 1.71 |
| Unsupported claims · sales brief% of products, lower is better 19 · 9 of 20 measured | 84.5% | 59.5% |
| Wrong attributes% of products, lower is better 16 · 12 of 20 measured | 21.4% | 19.1% |
| Wrong attributes · sales brief% of products, lower is better 13 · 16 of 20 measured | 19.6% | 20.8% |
| Wrong category or variant% of products, lower is better 4 · 12 of 20 measured | 1.8% | 3.0% |
| Wrong category or variant · sales brief% of products, lower is better 10 · 13 of 20 measured | 2.4% | 3.0% |
| DeckBench measured by Spring Prompt · 13 results | ||
| Caveat dropped% of tasks, lower is better 10 · 10 of 19 measured | 16.7% | 16.7% |
| Clean layout% of tasks, higher is better 6 · 1 of 19 measured | 66.7% | 100.0% |
| Design quality% of the maximum, higher is better 9 · 4 of 19 measured | 67.0% | 78.5% |
| Draft figure quoted% of tasks, lower is better 11 · 1 of 19 measured | 33.3% | 0.0% |
| Findings missing% of tasks, lower is better 1 · 1 of 19 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. 10 · 4 of 19 measured | 61.5% | 80.4% |
| Layout defects% of tasks, lower is better 6 · 1 of 19 measured | 33.3% | 0.0% |
| Misleading metric used% of tasks, lower is better 1 · 1 of 19 measured | 0.0% | 0.0% |
| Presentable decks% of tasks, higher is better 5 · 2 of 19 measured | 33.3% | 50.0% |
| Recommendation late or wrong% of tasks, lower is better 1 · 7 of 19 measured | 0.0% | 16.7% |
| Slides needing work% of tasks, lower is better 10 · 7 of 19 measured | 50.0% | 33.3% |
| Unsupported claims% of tasks, lower is better 10 · 1 of 19 measured | 16.7% | 0.0% |
| Unsupported numbers% of tasks, lower is better 1 · 1 of 19 measured | 0.0% | 0.0% |
| ROASBench measured by Spring Prompt · 6 results | ||
| Consistency scorescore out of 100, higher is better 13 · 9 of 19 measured | 19.3 | 28.5 |
| Contribution profitsimulated US dollars, higher is better 14 · 8 of 19 measured | $99,775 | $571,203 |
| Cost of a runUS dollars, lower is better 15 · 12 of 19 measured | $0.73 | $0.42 |
| Months over budgetmonths of 12, lower is better 17 · 1 of 19 measured | 2 | 0 |
| Planning scorescore out of 100, higher is better 9 · 2 of 19 measured | 54.8 | 55.1 |
| Return on ad spendprofit per $1 spent, higher is better 14 · 1 of 19 measured | $0.09 | $0.70 |
| Arena (formerly LMArena) reported by Arena (formerly LMArena) · 8 results | ||
| Confirmed task success · AgentIPS effect estimate, higher is better 20 · 5 of 48 measured | 0.05arena agent | 0.14max reasoning, arena agent |
| Creative writing · TextArena rating, higher is better 17 · 34 of 385 measured | 1,468 | 1,447xhigh reasoning |
| Expert prompts · TextArena rating, higher is better 20 · 1 of 338 measured | 1,513 | 1,556xhigh reasoning |
| Instruction following · TextArena rating, higher is better 25 · 15 of 387 measured | 1,474 | 1,484xhigh reasoning |
| Praise over complaint · AgentIPS effect estimate, higher is better 20 · 5 of 48 measured | 0.03arena agent | 0.26max reasoning, arena agent |
| Steerability · AgentIPS effect estimate, higher is better 18 · 8 of 48 measured | 0.03arena agent | 0.07max reasoning, arena agent |
| Tool grounding · AgentIPS effect estimate, higher is better 43 · 7 of 48 measured | -0.01arena agent | 0.00max reasoning, arena agent |
| Writing, literature and language · TextArena rating, higher is better 18 · 27 of 386 measured | 1,472 | 1,461xhigh reasoning |
| Artificial Analysis reported by Artificial Analysis · 4 results | ||
| Long-context reasoning (AA-LCR)% of questions, higher is better 35 · 19 of 233 measured | 80.3% | 82.7%max reasoning |
| Output speedtokens per second, higher is better 71 · 24 of 74 measured | 37.0 | 134max reasoning |
| SciCode% of problems, higher is better 36 · 4 of 113 measured | 52.1% | 61.0%max reasoning |
| Time to first answer tokenseconds, lower is better 74 · 17 of 74 measured | 55.8 | 0.76medium reasoning |
Compare them with other models
- Alibaba vs AnthropicEvery model from both labs, best against best
- Qwen3.8-Max (0902) vs Claude Opus 5.5Most intelligent at $3–10 per million tokens
- Qwen3.8-Max (0902) vs Claude Haiku 5.5Cheapest with similar intelligence
- Claude Sonnet 5.5 vs Claude Opus 5.5Most intelligent at $3–10 per million tokens
- Claude Sonnet 5.5 vs Muse Spark 1.3Just above overall
- Claude Sonnet 5.5 vs Gemini 3.8 FlashJust below overall
Quick answers
Is Qwen3.8-Max (0902) better than Claude Sonnet 5.5?
Pick Claude Sonnet 5.5: it scores higher (Spring Prompt overall 64 against 48) and costs less.
Which is cheaper, Qwen3.8-Max (0902) or Claude Sonnet 5.5?
Claude Sonnet 5.5 is cheaper to run: $0.017 against $0.026 per product listing on CatalogBench (33% less).
Which is faster, Qwen3.8-Max (0902) or Claude Sonnet 5.5?
Claude Sonnet 5.5 writes about 134 tokens a second on its usual API, against 37 for Qwen3.8-Max (0902) (Artificial Analysis).
Which is better for product listings, Qwen3.8-Max (0902) or Claude Sonnet 5.5?
Claude Sonnet 5.5 ranks 10 of 20 models on sale for turning a sparse product feed and photos into listings that can go live, against 17 for Qwen3.8-Max (0902).
Which is better for decks from an analysis, Qwen3.8-Max (0902) or Claude Sonnet 5.5?
Claude Sonnet 5.5 ranks 4 of 19 models on sale for turning a finished analysis into a deck you could present as it is, against 11 for Qwen3.8-Max (0902).
Which is better for marketing planning, Qwen3.8-Max (0902) or Claude Sonnet 5.5?
Claude Sonnet 5.5 ranks 4 of 19 models on sale for planning a year of ad spend without overspending, against 14 for Qwen3.8-Max (0902).
Which is better for agents and tool use, Qwen3.8-Max (0902) or Claude Sonnet 5.5?
Claude Sonnet 5.5 ranks 1 of 164 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 7 for Qwen3.8-Max (0902).
Run Qwen3.8-Max (0902) and Claude Sonnet 5.5 on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.