Compare / gpt-oss-20b vs o3-mini

gpt-oss-20b vs o3-mini

Which is cheaper, which is more capable, and which does better at each kind of work, from 18 results on 2 sources that measured both models.

OpenAIgpt-oss-20b Released 5 Aug 2025 · $0.030 in · $0.14 out
OpenAIo3-mini Released 31 Jan 2025 · Not on sale
gpt-oss-20b wins 52 too close to callo3-mini wins 11

Of the 18 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).

gpt-oss-20b or o3-mini? The short answer

Pick o3-mini: it scores higher (Intelligence Index 12.5 against 10.0) at about the same price.

  • o3-mini scores higher on the Artificial Analysis Intelligence Index: 12.5 against 10.0.
  • gpt-oss-20b is newer: released 5 Aug 2025, against 31 Jan 2025 for o3-mini.
  • On third-party benchmarks, on the results both models have: gpt-oss-20b does better for agents and tool use; o3-mini for reasoning and knowledge and writing.

Head to head

10.0
Intelligence Index
12.5

Every fact side by side

gpt-oss-20bo3-mini
Price per million tokens$0.030 in · $0.14 outNot on sale
Blended (3 in : 1 out)$0.058–
Cheapest host, blended$0.022 (PrimeIntellect, FP4)–
Intelligence Index10.012.5
Spring Prompt overall––
Output speed, tokens a second277–
Context window131,072 tokens–
Released5 Aug 2025 (newer)31 Jan 2025
WeightsOpenClosed
Developer based inthe USthe 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 130 of 159 144 of 160
Reasoning and knowledgeHard questions across science, maths and general knowledge 132 of 187 124 of 188
WritingWhat people prefer in blind comparisons, and judged writing quality 155 of 169 125 of 170
Sticking to the factsSummarising without inventing things, and factual answers not measured 127 of 140

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.

Measuregpt-oss-20bo3-mini
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 237 · 176 of 387 measured 1,318 1,364high reasoning
Business, management and finance · TextArena rating, higher is better 213 · 191 of 380 measured 1,323 1,349high reasoning
Creative writing · TextArena rating, higher is better 275 · 191 of 385 measured 1,240 1,312high reasoning
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 176 · 156 of 249 measured 10.0low reasoning 12.5
GPQA Diamond% of questions, higher is better 159 · 126 of 232 measured 68.8%high reasoning 77.3%high reasoning
Humanity's Last Exam% of questions, higher is better 153 · 147 of 250 measured 11.0%high reasoning 12.0%high reasoning
Show all 18 results (12 more)
Measuregpt-oss-20bo3-mini
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 220 · 157 of 338 measured 1,314 1,398high reasoning
Instruction following · TextArena rating, higher is better 257 · 165 of 387 measured 1,281 1,366high reasoning
Writing, literature and language · TextArena rating, higher is better 261 · 178 of 386 measured 1,275 1,341high reasoning
Artificial Analysis reported by Artificial Analysis · 9 results
Terminal-Bench 2.1% of tasks, higher is better 97 · 115 of 131 measured 13.9%high reasoning 4.5%high reasoning
Terminal-Bench 4.0% of tasks, higher is better 72 · 72 of 120 measured 0.0%high reasoning 0.0%high reasoning
τ-bench banking% of tasks, higher is better 90 · 106 of 121 measured 7.0%high reasoning 5.2%high reasoning
Artificial Analysis Coding Indexindex score, higher is better 100 · 106 of 130 measured 20.7high reasoning 16.3high reasoning
IFBench% of instructions, higher is better 69 · 66 of 195 measured 65.1%high reasoning 67.1%high reasoning
Long-context reasoning (AA-LCR)% of questions, higher is better 184 · 174 of 241 measured 34.7%high reasoning 43.0%high reasoning
SciCode% of problems, higher is better 90 · 79 of 121 measured 38.9%high reasoning 42.8%high reasoning
Terminal-Bench Hard% of tasks, higher is better 127 · 140 of 190 measured 10.6%high reasoning 6.8%
τ²-bench telecom% of tasks, higher is better 96 · 132 of 193 measured 60.2%high reasoning 31.3%high reasoning

Quick answers

Is gpt-oss-20b better than o3-mini?

Pick o3-mini: it scores higher (Intelligence Index 12.5 against 10.0) at about the same price.

Which is better for agents and tool use, gpt-oss-20b or o3-mini?

gpt-oss-20b ranks 130 of 159 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 144 for o3-mini.

Which is better for reasoning and knowledge, gpt-oss-20b or o3-mini?

o3-mini ranks 124 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, gpt-oss-20b or o3-mini?

o3-mini ranks 125 of 169 models on sale for what people prefer in blind comparisons, and judged writing quality, against 155 for gpt-oss-20b.

Run gpt-oss-20b and o3-mini 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.