ResearchLLM Verifications Leaderboard

Data as of · 78 models, 80 questions · reviewed by Suhas Bhairav · methodology

Head to head · tested 27 Oct 2026

gpt-oss-120b vs gpt-oss-20b: accuracy, speed and price

Similar accuracy; gpt-oss-20b costs less per question.

OpenAIgpt-oss-120b58.8% score · #41$0.00013 per question82 tok/sOpenAIgpt-oss-20b51.2% score · #46$0.000066 per question103 tok/s

Verdict

Built from the measurements below

  • Accuracy is too close to call: gpt-oss-120b scored 58.8% and gpt-oss-20b 51.2% on the same 80 questions, a gap within what chance alone could produce.
  • gpt-oss-20b is 1.9× cheaper per question as billed ($0.000066 vs $0.00013), reasoning tokens included.
  • gpt-oss-20b generates text 1.3× faster: 103 tokens per second against 82.
  • gpt-oss-120b starts answering sooner: 0.5 s to the first token, against 0.7 s for gpt-oss-20b.

Which should you choose?

Choose gpt-oss-120b for a quicker first token and more hosts to choose from on OpenRouter (23 against 10). Choose gpt-oss-20b for the lower cost per correct answer ($0.00013 against $0.00021) and faster streaming. The widest subject gap is in science: gpt-oss-120b answered 23 of 30 correctly, gpt-oss-20b 18 of 30. With so few questions per subject, treat this as a lead to test on your own prompts, not a result.

Accuracy gap +7.5 points (z = 0.95). A winner is called only when z is at least 1.96, the usual 95% threshold. Speed, latency and price need a 10% margin.

Head to head

Same questions, same prompt, same API

Measuregpt-oss-120bgpt-oss-20bEdge
Verification scoreHigher is better58.8%51.2%Too close to call
Cost per questionLower is better$0.00013$0.000066gpt-oss-20b
Output speedHigher is better82103gpt-oss-20b
Time to first tokenLower is better0.5 s0.7 sgpt-oss-120b
End-to-end response timeLower is better5.4 s4.0 sgpt-oss-20b
Tokens per questionLower is better546472gpt-oss-20b
Context windowHigher is better131k131kSimilar

Side by side

Bars are scaled to the larger of the two values. The winner on each measure is marked in bold.

Every measure

gpt-oss-120b first, gpt-oss-20b second

Verification scorehigher is better
gpt-oss-120b58.8%
gpt-oss-20b51.2%
Cost per questionlower is better
gpt-oss-120b$0.00013
gpt-oss-20b$0.000066
Output speedhigher is better
gpt-oss-120b82
gpt-oss-20b103
Time to first tokenlower is better
gpt-oss-120b0.5 s
gpt-oss-20b0.7 s
End-to-end response timelower is better
gpt-oss-120b5.4 s
gpt-oss-20b4.0 s
Tokens per questionlower is better
gpt-oss-120b546
gpt-oss-20b472
Context windowhigher is better
gpt-oss-120b131k
gpt-oss-20b131k

Score vs. cost per question

Both models in color, every other model in grey

Most attractive quadrant0255075100$0.00001$0.0001$0.001$0.01$0.10Cost per question (USD, log) →Verification score (%) →Qwen3.8 2.4T A95BCommand AReka Flash 3Gemma 4 26B A4BGemma 4 31BQwen3.5 397B A17BHermes 4 405BMuse WonderGrok 4.7o3gpt-oss-20bMinistral 3 3B 2512Nova Micro 1.0Phi 4Granite 4.2 8BHy3GLM 5.3 FlashSolar Pro 4Trinity Large ThinkingMiniMax M3Mistral Large 4Claude Opus 5.5GPT-6.1 SolGemini 4 ArgonGPT-5.4GPT-6 Lunagpt-oss-120bLlama 4 Maverick

Specifications and pricing

From OpenRouter's catalogue, plus the leaderboard's test dates

Specificationgpt-oss-120bgpt-oss-20b
CreatorOpenAIOpenAI
Released5 Aug 20255 Aug 2025
WeightsOpen weightsOpen weights
ReasoningYesYes
Inputtexttext
Outputtexttext
Context window131k tokens131k tokens
Knowledge cutoff30 Jun 202430 Jun 2024
List price in / out per 1M$0.15 / $0.6$0.018 / $0.09
Blended price per 1M (3:1)$0.26$0.04
Cheapest hostNear AI · $0.03 / $0.15Darkbloom · $0.018 / $0.09
Hosts on OpenRouter2310
Last tested25 Oct 202626 Oct 2026

Compare any two models →

gpt-oss-120b vs gpt-oss-20b: frequently asked questions

Is gpt-oss-120b better than gpt-oss-20b?
On accuracy they are statistically tied: gpt-oss-120b scored 58.8% and gpt-oss-20b 51.2% on the same 80 questions from open datasets. Similar accuracy; gpt-oss-20b costs less per question.
Which is cheaper, gpt-oss-120b or gpt-oss-20b?
gpt-oss-20b is cheaper: $0.000066 per question as billed, against $0.00013. List prices: gpt-oss-120b $0.15 in / $0.6 out, gpt-oss-20b $0.018 in / $0.09 out per million tokens.
Which is faster, gpt-oss-120b or gpt-oss-20b?
gpt-oss-20b is faster at 103 output tokens per second, against 82. Time to first token: 0.5 s for gpt-oss-120b, 0.7 s for gpt-oss-20b.
Which has the larger context window, gpt-oss-120b or gpt-oss-20b?
Both accept about the same amount of context: 131k and 131k tokens.