Compare / DeepSeek-R1 vs gpt-oss-20b

DeepSeek-R1 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.

DeepSeekDeepSeek-R1 Released 20 Jan 2025 · Not on sale
OpenAIgpt-oss-20b Released 5 Aug 2025 · $0.030 in · $0.14 out
DeepSeek-R1 wins 131 too close to callgpt-oss-20b wins 7

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 19 November 2026 (catalogue release 2026-11-19-3608133e8331). See also DeepSeek vs OpenAI, every model from both labs.

DeepSeek-R1 or gpt-oss-20b? The short answer

They score and cost about the same: pick on the results below that match your work.

  • DeepSeek-R1 scores higher on the Artificial Analysis Intelligence Index: 11.4 against 10.0.
  • gpt-oss-20b is newer: released 5 Aug 2025, against 20 Jan 2025 for DeepSeek-R1.
  • On third-party benchmarks, on the results both models have: DeepSeek-R1 does better for reasoning and knowledge and writing.

Head to head

11.4
Intelligence Index
10.0

Every fact side by side

DeepSeek-R1gpt-oss-20b
Price per million tokensNot on sale$0.030 in · $0.14 out
Blended (3 in : 1 out)–$0.058
Cheapest host, blended–$0.022 (PrimeIntellect, FP4)
Intelligence Index11.410.0
Spring Prompt overall––
Output speed, tokens a second–277
Context window–131,072 tokens
Released20 Jan 20255 Aug 2025 (newer)
WeightsClosedOpen
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 129 of 160 130 of 159
Reasoning and knowledgeHard questions across science, maths and general knowledge 132 of 188 132 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality 90 of 170 155 of 169
Sticking to the factsSummarising without inventing things, and factual answers 104 of 140 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.

MeasureDeepSeek-R1gpt-oss-20b
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 141 · 237 of 387 measured 1,398 1,318
Business, management and finance · TextArena rating, higher is better 151 · 213 of 380 measured 1,388 1,323
Creative writing · TextArena rating, higher is better 125 · 275 of 385 measured 1,374 1,240
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 163 · 176 of 249 measured 11.4 10.0low reasoning
GPQA Diamond% of questions, higher is better 155 · 159 of 232 measured 70.8% 68.8%high reasoning
Humanity's Last Exam% of questions, higher is better 170 · 153 of 250 measured 8.5% 11.0%high reasoning
UGI Leaderboard reported by UGI Leaderboard · 3 results
Writing score · UGIscore out of 100, higher is better 74 · 182 of 230 measured 51.8 24.6low reasoning
Requested-length error · UGI% off the requested word count, lower is better 91 · 221 of 230 measured 18.0% 98.0%low reasoning
Style adherence · UGIscore from 0 to 1, higher is better 203 · 187 of 230 measured 0.30 0.31medium reasoning
Show all 21 results (12 more)
MeasureDeepSeek-R1gpt-oss-20b
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 151 · 220 of 338 measured 1,402 1,314
Instruction following · TextArena rating, higher is better 128 · 257 of 387 measured 1,397 1,281
Writing, literature and language · TextArena rating, higher is better 130 · 261 of 386 measured 1,382 1,275
Artificial Analysis reported by Artificial Analysis · 9 results
Terminal-Bench 2.1% of tasks, higher is better 93 · 97 of 131 measured 19.1% 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 94 · 90 of 121 measured 6.4% 7.0%high reasoning
Artificial Analysis Coding Indexindex score, higher is better 91 · 100 of 130 measured 24.6 20.7high reasoning
IFBench% of instructions, higher is better 141 · 69 of 195 measured 39.0% 65.1%high reasoning
Long-context reasoning (AA-LCR)% of questions, higher is better 148 · 184 of 241 measured 57.7% 34.7%high reasoning
SciCode% of problems, higher is better 92 · 90 of 121 measured 38.3% 38.9%high reasoning
Terminal-Bench Hard% of tasks, higher is better 147 · 127 of 190 measured 6.1% 10.6%high reasoning
τ²-bench telecom% of tasks, higher is better 182 · 96 of 193 measured 11.4% 60.2%high reasoning

Quick answers

Is DeepSeek-R1 better than gpt-oss-20b?

They score and cost about the same: pick on the results below that match your work.

Which is better for agents and tool use, DeepSeek-R1 or gpt-oss-20b?

DeepSeek-R1 ranks 129 of 160 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 130 for gpt-oss-20b.

Which is better for writing, DeepSeek-R1 or gpt-oss-20b?

DeepSeek-R1 ranks 90 of 170 models on sale for what people prefer in blind comparisons, and judged writing quality, against 155 for gpt-oss-20b.

Run DeepSeek-R1 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.