Compare / Qwen3-32B vs GPT-5

Qwen3-32B vs GPT-5

Which is cheaper, which is more capable, and which does better at each kind of work, from 21 results on 4 sources that measured both models.

AlibabaQwen3-32B Released 28 Apr 2025 · $0.14 in · $0.40 out
OpenAIGPT-5 Released 7 Aug 2025 · $1.25 in · $10.00 out
Qwen3-32B wins 21 too close to callGPT-5 wins 18

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 9 October 2026 (catalogue release 2026-10-09-dd620e778ac5). See also Alibaba vs OpenAI, every model from both labs.

Qwen3-32B or GPT-5? The short answer

Pick GPT-5 if quality matters most: it scores higher (Intelligence Index 23.0 against 8.6). Pick Qwen3-32B if cost matters more: it costs 17 times less per token.

  • Qwen3-32B is cheaper: $0.21 against $3.44 per million tokens, 17 times less.
  • GPT-5 scores higher on the Artificial Analysis Intelligence Index: 23.0 against 8.6.
  • GPT-5 is newer: released 7 Aug 2025, against 28 Apr 2025 for Qwen3-32B.
  • On third-party benchmarks, on the results both models have: Qwen3-32B does better for sticking to the facts; GPT-5 for agents and tool use, reasoning and knowledge and writing.

Head to head

8.6
Intelligence Index
23.0
$0.21
Price per million tokens, blendedlower is better
$3.44
131,072
Context window, tokens
400,000

Every fact side by side

Qwen3-32BGPT-5
Price per million tokens$0.14 in · $0.40 out$1.25 in · $10.00 out
Blended (3 in : 1 out)$0.21$3.44
Cheapest host, blended$0.13 (DeepInfra, FP8)$3.44 (Azure)
Intelligence Index8.623.0
Spring Prompt overall––
Output speed, tokens a second––
Context window131,072 tokens400,000 tokens
Released28 Apr 20257 Aug 2025 (newer)
WeightsOpenClosed
Developer based inChinathe 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.

Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 113 of 164 78 of 164
Professional workReal tasks from banking, consulting and law, business documents and freelance projects not measured 44 of 51
Reasoning and knowledgeHard questions across science, maths and general knowledge 148 of 198 76 of 198
WritingWhat people prefer in blind comparisons, and judged writing quality 139 of 188 70 of 188
Sticking to the factsSummarising without inventing things, and factual answers 22 of 152 99 of 152

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.

MeasureQwen3-32BGPT-5
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 195 · 90 of 387 measured 1,347 1,435high reasoning
Business, management and finance · TextArena rating, higher is better 203 · 118 of 380 measured 1,340 1,413high reasoning
Creative writing · TextArena rating, higher is better 205 · 128 of 385 measured 1,304 1,373high reasoning
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 188 · 85 of 243 measured 8.6thinking reasoning 23.0high reasoning
GPQA Diamond% of questions, higher is better 167 · 77 of 231 measured 66.8%thinking reasoning 85.4%high reasoning
Humanity's Last Exam% of questions, higher is better 168 · 79 of 242 measured 7.4%thinking reasoning 28.5%high reasoning
UGI Leaderboard reported by UGI Leaderboard · 3 results
Writing score · UGIscore out of 100, higher is better 134 · 68 of 230 measured 33.0no reasoning 53.5high reasoning
Requested-length error · UGI% off the requested word count, lower is better 99 · 207 of 230 measured 19.0%no reasoning 58.0%high reasoning
Style adherence · UGIscore from 0 to 1, higher is better 175 · 20 of 230 measured 0.32no reasoning 0.39high reasoning
Vectara Hallucination Leaderboard reported by Vectara Hallucination Leaderboard · 2 results
Hallucination rate · Vectara% of summaries, lower is better 24 · 91 of 104 measured 5.9% 14.7%minimal reasoning
Answer rate · Vectara% of documents, higher is better 17 · 17 of 104 measured 99.9% 99.9%high reasoning
Show all 21 results (10 more)
MeasureQwen3-32BGPT-5
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 159 · 90 of 338 measured 1,398 1,459high reasoning
Instruction following · TextArena rating, higher is better 198 · 110 of 387 measured 1,331 1,410high reasoning
Writing, literature and language · TextArena rating, higher is better 217 · 113 of 386 measured 1,308 1,396high reasoning
Artificial Analysis reported by Artificial Analysis · 7 results
Terminal-Bench 2.1% of tasks, higher is better 111 · 78 of 130 measured 5.2%thinking reasoning 35.2%high reasoning
τ-bench banking% of tasks, higher is better 102 · 45 of 121 measured 5.4%thinking reasoning 22.1%high reasoning
Artificial Analysis Coding Indexindex score, higher is better 111 · 76 of 131 measured 15.3thinking reasoning 37.8high reasoning
IFBench% of instructions, higher is better 155 · 36 of 195 measured 36.3%thinking reasoning 73.1%high reasoning
Long-context reasoning (AA-LCR)% of questions, higher is better 222 · 58 of 233 measured 0.0%no reasoning 78.2%high reasoning
Terminal-Bench Hard% of tasks, higher is better 169 · 39 of 190 measured 3.0%thinking reasoning 37.9%medium reasoning
τ²-bench telecom% of tasks, higher is better 135 · 54 of 193 measured 29.8%thinking reasoning 86.6%medium reasoning

Quick answers

Is Qwen3-32B better than GPT-5?

Pick GPT-5 if quality matters most: it scores higher (Intelligence Index 23.0 against 8.6). Pick Qwen3-32B if cost matters more: it costs 17 times less per token.

Which is cheaper, Qwen3-32B or GPT-5?

Qwen3-32B is cheaper: $0.21 against $3.44 per million tokens, 17 times less.

Which is better for agents and tool use, Qwen3-32B or GPT-5?

GPT-5 ranks 78 of 164 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 113 for Qwen3-32B.

Which is better for reasoning and knowledge, Qwen3-32B or GPT-5?

GPT-5 ranks 76 of 198 models on sale for hard questions across science, maths and general knowledge, against 148 for Qwen3-32B.

Which is better for writing, Qwen3-32B or GPT-5?

GPT-5 ranks 70 of 188 models on sale for what people prefer in blind comparisons, and judged writing quality, against 139 for Qwen3-32B.

Which is better for sticking to the facts, Qwen3-32B or GPT-5?

Qwen3-32B ranks 22 of 152 models on sale for summarising without inventing things, and factual answers, against 99 for GPT-5.

Which has the bigger context window, Qwen3-32B or GPT-5?

GPT-5: 400,000 tokens, against 131,072 for Qwen3-32B.

Run Qwen3-32B and GPT-5 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