Compare / Qwen3-235B-A22B-Thinking-2507 vs gpt-oss-20b

Qwen3-235B-A22B-Thinking-2507 vs gpt-oss-20b

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.

AlibabaQwen3-235B-A22B-Thinking-2507 Released 25 Jul 2025 · $0.45 in · $3.50 out
OpenAIgpt-oss-20b Released 5 Aug 2025 · $0.030 in · $0.14 out
Qwen3-235B-A22B-Thinking-2507 wins 161 too close to callgpt-oss-20b wins 4

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 16 November 2026 (catalogue release 2026-11-16-49bdfa045dee). See also Alibaba vs OpenAI, every model from both labs.

Qwen3-235B-A22B-Thinking-2507 or gpt-oss-20b? The short answer

Pick Qwen3-235B-A22B-Thinking-2507 if quality matters most: it scores higher (Intelligence Index 12.7 against 10.0). Pick gpt-oss-20b if cost matters more: it costs 21 times less per token.

  • gpt-oss-20b is cheaper: $0.058 against $1.21 per million tokens, 21 times less.
  • Qwen3-235B-A22B-Thinking-2507 scores higher on the Artificial Analysis Intelligence Index: 12.7 against 10.0.
  • gpt-oss-20b is newer: released 5 Aug 2025, against 25 Jul 2025 for Qwen3-235B-A22B-Thinking-2507.
  • On third-party benchmarks, on the results both models have: Qwen3-235B-A22B-Thinking-2507 does better for reasoning and knowledge and writing.

Head to head

12.7
Intelligence Index
10.0
$1.21
Price per million tokens, blendedlower is better
$0.058
128,000
Context window, tokens
131,072

Every fact side by side

Qwen3-235B-A22B-Thinking-2507gpt-oss-20b
Price per million tokens$0.45 in · $3.50 out$0.030 in · $0.14 out
Blended (3 in : 1 out)$1.21$0.058
Cheapest host, blended–$0.022 (PrimeIntellect, FP4)
Intelligence Index12.710.0
Spring Prompt overall––
Output speed, tokens a second–277
Context window128,000 tokens131,072 tokens
Released25 Jul 20255 Aug 2025 (newer)
WeightsOpenOpen
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 139 of 159 130 of 159
Reasoning and knowledgeHard questions across science, maths and general knowledge 119 of 187 132 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality 119 of 169 155 of 169
Sticking to the factsSummarising without inventing things, and factual answers 83 of 139 not measured

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.

MeasureQwen3-235B-A22B-Thinking-2507gpt-oss-20b
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 139 · 237 of 387 measured 1,400 1,318
Business, management and finance · TextArena rating, higher is better 133 · 213 of 380 measured 1,404 1,323
Creative writing · TextArena rating, higher is better 133 · 275 of 385 measured 1,371 1,240
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 152 · 176 of 249 measured 12.7 10.0low reasoning
GPQA Diamond% of questions, higher is better 115 · 159 of 232 measured 79.0% 68.8%high reasoning
Humanity's Last Exam% of questions, higher is better 130 · 153 of 250 measured 15.9% 11.0%high reasoning
UGI Leaderboard reported by UGI Leaderboard · 3 results
Writing score · UGIscore out of 100, higher is better 139 · 182 of 230 measured 32.4 24.6low reasoning
Requested-length error · UGI% off the requested word count, lower is better 166 · 221 of 230 measured 30.0% 98.0%low reasoning
Style adherence · UGIscore from 0 to 1, higher is better 194 · 187 of 230 measured 0.31 0.31medium reasoning
Show all 21 results (12 more)
MeasureQwen3-235B-A22B-Thinking-2507gpt-oss-20b
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 85 · 220 of 338 measured 1,466 1,314
Instruction following · TextArena rating, higher is better 142 · 257 of 387 measured 1,385 1,281
Writing, literature and language · TextArena rating, higher is better 139 · 261 of 386 measured 1,377 1,275
Artificial Analysis reported by Artificial Analysis · 9 results
Terminal-Bench 2.1% of tasks, higher is better 102 · 97 of 131 measured 12.0% 13.9%high reasoning
Terminal-Bench 4.0% of tasks, higher is better 72 · 72 of 120 measured 0.0% 0.0%high reasoning
τ-bench banking% of tasks, higher is better 87 · 90 of 121 measured 7.8% 7.0%high reasoning
Artificial Analysis Coding Indexindex score, higher is better 97 · 100 of 130 measured 22.1 20.7high reasoning
IFBench% of instructions, higher is better 101 · 69 of 195 measured 51.2% 65.1%high reasoning
Long-context reasoning (AA-LCR)% of questions, higher is better 105 · 184 of 241 measured 72.0% 34.7%high reasoning
SciCode% of problems, higher is better 82 · 90 of 121 measured 41.4% 38.9%high reasoning
Terminal-Bench Hard% of tasks, higher is better 116 · 127 of 190 measured 13.6% 10.6%high reasoning
τ²-bench telecom% of tasks, higher is better 101 · 96 of 193 measured 53.2% 60.2%high reasoning

Quick answers

Is Qwen3-235B-A22B-Thinking-2507 better than gpt-oss-20b?

Pick Qwen3-235B-A22B-Thinking-2507 if quality matters most: it scores higher (Intelligence Index 12.7 against 10.0). Pick gpt-oss-20b if cost matters more: it costs 21 times less per token.

Which is cheaper, Qwen3-235B-A22B-Thinking-2507 or gpt-oss-20b?

gpt-oss-20b is cheaper: $0.058 against $1.21 per million tokens, 21 times less.

Which is better for agents and tool use, Qwen3-235B-A22B-Thinking-2507 or gpt-oss-20b?

gpt-oss-20b ranks 130 of 159 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 139 for Qwen3-235B-A22B-Thinking-2507.

Which is better for reasoning and knowledge, Qwen3-235B-A22B-Thinking-2507 or gpt-oss-20b?

Qwen3-235B-A22B-Thinking-2507 ranks 119 of 187 models on sale for hard questions across science, maths and general knowledge, against 132 for gpt-oss-20b.

Which is better for writing, Qwen3-235B-A22B-Thinking-2507 or gpt-oss-20b?

Qwen3-235B-A22B-Thinking-2507 ranks 119 of 169 models on sale for what people prefer in blind comparisons, and judged writing quality, against 155 for gpt-oss-20b.

Which has the bigger context window, Qwen3-235B-A22B-Thinking-2507 or gpt-oss-20b?

gpt-oss-20b: 131,072 tokens, against 128,000 for Qwen3-235B-A22B-Thinking-2507.

Run Qwen3-235B-A22B-Thinking-2507 and gpt-oss-20b 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.