Skip to main content

Model profile

GPT-OSS 20B

OpenAIEstablishedReleased Aug 5, 2025
Data verified
Overall Score
42.74Public #159 of 200Verified #75 of 99
Arena Elo
1317
Eligible category ranks
2of 8
Price (1M tokens)
$0 in / $0 out
API pricing
Speed
313tok/s
Context
128K

Evidence coverage

18 of 321 tracked benchmarks are published. 1 is verified and 17 provisional. 6 of 8 categories are measured.

Updated July 20, 2026Methodology
Published / tracked
18 / 321
Verified
1
Provisional
17
Categories with evidence
6 / 8

Evidence by category

  • Agentic5 benchmarks
    Reported
  • Coding3 benchmarks
    Mixed evidence
  • Reasoning2 benchmarks
    Reported
  • Knowledge6 benchmarks
    Reported
  • Math0 benchmarks
    Not measured
  • Multilingual0 benchmarks
    Not measured
  • Multimodal1 benchmark
    Reported
  • Inst. Following1 benchmark
    Reported
Open WeightSelf-hostNon-Reasoning
Confidence:
Low
mini

GPT-OSS 20B ranks #159 out of 200 models on the public leaderboard with an overall score of 42.74/100. It also ranks #75 out of 99 on the verified leaderboard. While not a frontier model, it offers specific advantages depending on the use case.

GPT-OSS 20B is a open weight model with a 128K token context window. It processes queries without explicit chain-of-thought reasoning, offering faster response times and lower token usage.

GPT-OSS 20B sits inside the GPT-OSS family alongside GPT-OSS 120B. This profile currently has 18 of 321 tracked benchmarks. BenchLM only exposes non-generated benchmark rows publicly, so missing categories stay blank until a sourced evaluation is available.

Its strongest category is Agentic (#101), while its weakest is Coding (#104). This performance profile makes it particularly useful for coding agents, browser research, and computer-use workflows.

Peer position

Exact provisional scores and ranks for the closest listed peers. A score can appear before a model clears the evidence threshold for a rank, so equal scores can have different rank states.

Range 42.0643.2

  1. Gemma 4 E4B
    Google
    #15543.2
    Gemma 4 E4B is #155 with a score of 43.2.
    Compare
  2. Mistral Medium 3
    Mistral
    #15643.2
    Mistral Medium 3 is #156 with a score of 43.2.
    Compare
  3. Sarvam 105B
    Sarvam
    #15742.97
    Sarvam 105B is #157 with a score of 42.97.
    Compare
  4. Claude 4 Sonnet
    Anthropic
    #15842.79
    Claude 4 Sonnet is #158 with a score of 42.79.
    Compare
  5. GPT-OSS 20BCurrent model
    OpenAI
    #15942.74
    GPT-OSS 20B is #159 with a score of 42.74.
  6. DeepSeek R1 Distill Qwen 32B
    DeepSeek
    #16042.58
    DeepSeek R1 Distill Qwen 32B is #160 with a score of 42.58.
    Compare
  7. GPT-4.1 nano
    OpenAI
    #16142.06
    GPT-4.1 nano is #161 with a score of 42.06.
    Compare

Category percentile

More

Relative position among models eligible for each sourced category. A higher percentile means a stronger position within that category's ranked cohort; 100 is highest.

  1. Agentic15%
    Eligible cohort rank #101 of 119Category score 37.4
  2. Coding15%
    Eligible cohort rank #104 of 122Category score 40.7

Category evidence

Scores and ranks appear only where this model has published benchmark evidence. Categories without displayable source records remain not measured.

Category scores, ranks, weighting, benchmark coverage, and evidence status
CategoryScore
AgenticRank #101 of 119Percentile 15thWeight 22%5 benchmarksReported37.4
CodingRank #104 of 122Percentile 15thWeight 20%3 benchmarksMixed sources40.7
ReasoningRank Not rankedWeight 17%2 benchmarksReported48.0
KnowledgeRank Not rankedWeight 12%6 benchmarksReported23.6
MathWeight 5%0 benchmarksNot measuredNot measured
MultilingualWeight 7%0 benchmarksNot measuredNot measured
MultimodalRank Not rankedWeight 12%1 benchmarkReported35.4
Inst. FollowingRank Not rankedWeight 5%1 benchmarkReported67.0

Chatbot Arena performance

Scroll horizontally to inspect confidence intervals and vote counts.

Chatbot Arena Elo, confidence interval, and vote count by evaluation view
ViewEloConfidence intervalVotes
Text Overall1317±6.410,622
Coding1369±12.92,168
Math1336±22.0679
Instruction Following1281±12.12,554
Creative Writing1238±18.31,243
Multi-turn1292±14.31,836
Hard Prompts1323±8.84,797
Hard Prompts (English)1339±12.32,447
Longer Query1301±13.42,090

Benchmark Details

Rows below have a displayable published verification record. Each source link and provenance note remains in the page HTML while its category is closed. Source-unverified manual rows and generated rows stay hidden.

Agentic5 benchmarks
AA Agentic IndexReported

Artificial Analysis Agentic Index

3.1%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
APEX-Agents-AAReported
0.7%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
τ²-bench resultsReported

τ²-Bench Tool-Agent-User Evaluation

60.2%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
GDPval-AAReported

GDPval-AA normalized

3.0%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
GDPval-AAReported
559Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Coding3 benchmarks
React Native EvalsBenchmark exact
71%Display only
Source: React Native Evals leaderboardProvenance: React Native Evals reports this exact overall score for GPT OSS 20B in the public dashboard run finished on 2026-04-28.
AA Coding IndexReported

Artificial Analysis Coding Index

20.7%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-SciCodeReported

Artificial Analysis SciCode

34.4%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Reasoning2 benchmarks
AA-LCRReported

Artificial Analysis Long Context Reasoning

30.7%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
CritPtReported

Critical Physics Tasks

1.4%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Knowledge6 benchmarks
Artificial Analysis Intelligence IndexReported
14.9%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-GPQA DiamondReported

Artificial Analysis GPQA Diamond

68.8%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-HLEReported

Artificial Analysis Humanity's Last Exam

9.8%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-Omniscience IndexReported

Artificial Analysis Omniscience Index

-63.9%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-Omniscience AccuracyReported

Artificial Analysis Omniscience Accuracy

15.5%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-Omniscience Hallucination RateReported

Artificial Analysis Omniscience Hallucination Rate

94.1%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Multimodal1 benchmark
Design Arena WebsiteReported

Design Arena Website Elo

885Display only
Source: OpenRouter model benchmarksProvenance: Display-only Design Arena Website Elo synced from OpenRouter model benchmark metadata. It is excluded from BenchLM weighted scoring.
Inst. Following1 benchmark
AA-IFBenchReported

Artificial Analysis IFBench

65.1%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.

GPT-OSS Family

Mini

Canonical Entry

GPT-OSS 120B

Frequently Asked Questions

How does GPT-OSS 20B perform overall in AI benchmarks?

GPT-OSS 20B has 18 published benchmark scores on BenchLM, but it does not yet have enough non-generated coverage to receive a global overall rank.

Is GPT-OSS 20B good for knowledge and understanding?

GPT-OSS 20B has visible benchmark coverage in knowledge and understanding, but BenchLM does not currently assign it a global category rank there.

Is GPT-OSS 20B good for coding and programming?

GPT-OSS 20B ranks #104 out of 122 models in coding and programming benchmarks with an average score of 40.7. There are stronger options in this category.

Is GPT-OSS 20B good for reasoning and logic?

GPT-OSS 20B has visible benchmark coverage in reasoning and logic, but BenchLM does not currently assign it a global category rank there.

Is GPT-OSS 20B good for agentic tool use and computer tasks?

GPT-OSS 20B ranks #101 out of 119 models in agentic tool use and computer tasks benchmarks with an average score of 37.4. There are stronger options in this category.

Is GPT-OSS 20B good for multimodal and grounded tasks?

GPT-OSS 20B has visible benchmark coverage in multimodal and grounded tasks, but BenchLM does not currently assign it a global category rank there.

Is GPT-OSS 20B good for instruction following?

GPT-OSS 20B has visible benchmark coverage in instruction following, but BenchLM does not currently assign it a global category rank there.

Is GPT-OSS 20B open source?

Yes, GPT-OSS 20B is an open weight model created by OpenAI, meaning it can be downloaded and run locally or fine-tuned for specific use cases.

Which sibling models are related to GPT-OSS 20B?

GPT-OSS 20B belongs to the GPT-OSS family. Related variants on BenchLM include GPT-OSS 120B.

Does GPT-OSS 20B have full benchmark coverage on BenchLM?

Not yet. GPT-OSS 20B currently has 18 published benchmark scores out of the 321 benchmarks BenchLM tracks. BenchLM only exposes non-generated public benchmark rows, so missing categories stay blank until a sourced evaluation is available.

What is the context window size of GPT-OSS 20B?

GPT-OSS 20B has a published context window of 128K, which determines how much text it can process in a single interaction.

Last updated: July 20, 2026 · Runtime metrics stay blank until BenchLM has a sourced snapshot.

Choose with this week’s evidence

Join 2,000+ readers for ranking moves, new releases, pricing changes, and the evidence behind them.

Free. One email per week.