Compare / Qwen3.5-397B-A17B vs Qwen3-Next-80B-A3B-Thinking

Qwen3.5-397B-A17B vs Qwen3-Next-80B-A3B-Thinking

Which is cheaper, which is more capable, and which does better at each kind of work, from 21 results on 3 sources that measured both models.

AlibabaQwen3.5-397B-A17B Released 13 Feb 2026 · $0.55 in · $3.50 out
AlibabaQwen3-Next-80B-A3B-Thinking Released 11 Sep 2025 · $0.15 in · $1.20 out
Qwen3.5-397B-A17B wins 20Qwen3-Next-80B-A3B-Thinking wins 1

Of the 21 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.

Results as of 14 October 2026 (catalogue release 2026-10-14-7d7cf32c3f6d).

Qwen3.5-397B-A17B or Qwen3-Next-80B-A3B-Thinking? The short answer

Pick Qwen3.5-397B-A17B if quality matters most: it scores higher (Intelligence Index 21.4 against 11.2). Pick Qwen3-Next-80B-A3B-Thinking if cost matters more: it costs 3.1 times less per token. Qwen3-Next-80B-A3B-Thinking is also faster.

  • Qwen3-Next-80B-A3B-Thinking is cheaper: $0.41 against $1.29 per million tokens, 3.1 times less.
  • Qwen3.5-397B-A17B scores higher on the Artificial Analysis Intelligence Index: 21.4 against 11.2.
  • Qwen3-Next-80B-A3B-Thinking writes faster: 174 tokens a second against 57.
  • Qwen3.5-397B-A17B is newer: released 13 Feb 2026, against 11 Sep 2025 for Qwen3-Next-80B-A3B-Thinking.
  • On third-party benchmarks, on the results both models have: Qwen3.5-397B-A17B does better for agents and tool use, reasoning and knowledge and writing.

Head to head

21.4
Intelligence Index
11.2
$1.29
Price per million tokens, blendedlower is better
$0.41
57
Output speed, tokens a second
174
262,144
Context window, tokens
262,144

Every fact side by side

Qwen3.5-397B-A17BQwen3-Next-80B-A3B-Thinking
Price per million tokens$0.55 in · $3.50 out$0.15 in · $1.20 out
Blended (3 in : 1 out)$1.29$0.41
Cheapest host, blended$0.88 (Alibaba)$0.41 (Alibaba)
Intelligence Index21.411.2
Spring Prompt overall––
Output speed, tokens a second57174
Context window262,144 tokens262,144 tokens
Released13 Feb 2026 (newer)11 Sep 2025
WeightsOpenOpen
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.

Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 71 of 166 142 of 166
Professional workReal tasks from banking, consulting and law, business documents and freelance projects 48 of 51 not measured
Reasoning and knowledgeHard questions across science, maths and general knowledge 75 of 200 128 of 200
WritingWhat people prefer in blind comparisons, and judged writing quality 70 of 189 144 of 189
Sticking to the factsSummarising without inventing things, and factual answers 76 of 153 64 of 153

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.

MeasureQwen3.5-397B-A17BQwen3-Next-80B-A3B-Thinking
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 80 · 170 of 387 measured 1,442 1,369
Business, management and finance · TextArena rating, higher is better 73 · 173 of 380 measured 1,446 1,367
Creative writing · TextArena rating, higher is better 86 · 180 of 385 measured 1,405 1,326
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 98 · 162 of 244 measured 21.4no reasoning 11.2
GPQA Diamond% of questions, higher is better 53 · 134 of 231 measured 89.3% 75.9%
Humanity's Last Exam% of questions, higher is better 78 · 137 of 243 measured 29.0% 12.6%
UGI Leaderboard reported by UGI Leaderboard · 3 results
Writing score · UGIscore out of 100, higher is better 77 · 167 of 230 measured 50.1no reasoning 27.2
Requested-length error · UGI% off the requested word count, lower is better 134 · 211 of 230 measured 23.0%no reasoning 67.0%
Style adherence · UGIscore from 0 to 1, higher is better 164 · 220 of 230 measured 0.33no reasoning 0.29
Show all 21 results (12 more)
MeasureQwen3.5-397B-A17BQwen3-Next-80B-A3B-Thinking
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 67 · 165 of 338 measured 1,477 1,393
Instruction following · TextArena rating, higher is better 78 · 171 of 387 measured 1,434 1,359
Writing, literature and language · TextArena rating, higher is better 80 · 177 of 386 measured 1,421 1,341
Artificial Analysis reported by Artificial Analysis · 9 results
Terminal-Bench 2.1% of tasks, higher is better 61 · 107 of 130 measured 51.3% 6.7%
τ-bench banking% of tasks, higher is better 64 · 95 of 121 measured 13.4% 6.2%
Artificial Analysis Coding Indexindex score, higher is better 61 · 106 of 131 measured 48.2 17.4
IFBench% of instructions, higher is better 8 · 79 of 195 measured 78.8% 60.7%
Long-context reasoning (AA-LCR)% of questions, higher is better 67 · 132 of 234 measured 77.3% 63.7%
Output speedtokens per second, higher is better 58 · 16 of 73 measured 57.1no reasoning 174
Terminal-Bench Hard% of tasks, higher is better 34 · 129 of 190 measured 40.9% 9.8%
Time to first answer tokenseconds, lower is better 35 · 45 of 73 measured 1.64no reasoning 12.0
τ²-bench telecom% of tasks, higher is better 16 · 113 of 193 measured 95.6% 41.5%

Quick answers

Is Qwen3.5-397B-A17B better than Qwen3-Next-80B-A3B-Thinking?

Pick Qwen3.5-397B-A17B if quality matters most: it scores higher (Intelligence Index 21.4 against 11.2). Pick Qwen3-Next-80B-A3B-Thinking if cost matters more: it costs 3.1 times less per token. Qwen3-Next-80B-A3B-Thinking is also faster.

Which is cheaper, Qwen3.5-397B-A17B or Qwen3-Next-80B-A3B-Thinking?

Qwen3-Next-80B-A3B-Thinking is cheaper: $0.41 against $1.29 per million tokens, 3.1 times less.

Which is faster, Qwen3.5-397B-A17B or Qwen3-Next-80B-A3B-Thinking?

Qwen3-Next-80B-A3B-Thinking writes about 174 tokens a second on its usual API, against 57 for Qwen3.5-397B-A17B (Artificial Analysis).

Which is better for agents and tool use, Qwen3.5-397B-A17B or Qwen3-Next-80B-A3B-Thinking?

Qwen3.5-397B-A17B ranks 71 of 166 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 142 for Qwen3-Next-80B-A3B-Thinking.

Which is better for reasoning and knowledge, Qwen3.5-397B-A17B or Qwen3-Next-80B-A3B-Thinking?

Qwen3.5-397B-A17B ranks 75 of 200 models on sale for hard questions across science, maths and general knowledge, against 128 for Qwen3-Next-80B-A3B-Thinking.

Which is better for writing, Qwen3.5-397B-A17B or Qwen3-Next-80B-A3B-Thinking?

Qwen3.5-397B-A17B ranks 70 of 189 models on sale for what people prefer in blind comparisons, and judged writing quality, against 144 for Qwen3-Next-80B-A3B-Thinking.

Which is better for sticking to the facts, Qwen3.5-397B-A17B or Qwen3-Next-80B-A3B-Thinking?

Qwen3-Next-80B-A3B-Thinking ranks 64 of 153 models on sale for summarising without inventing things, and factual answers, against 76 for Qwen3.5-397B-A17B.

Run Qwen3.5-397B-A17B and Qwen3-Next-80B-A3B-Thinking on your own prompt

Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.

Try the playgroundTalk to us