Compare · ModelsLive · 2 picked · head to head
Claude 3.5 Sonnet vs gpt-oss-20b
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
gpt-oss-20b wins on 6/10 benchmarks
gpt-oss-20b wins 6 of 10 shared benchmarks. Leads in general · knowledge · math.
Category leads
arena·Claude 3.5 Sonnetgeneral·gpt-oss-20bknowledge·gpt-oss-20blanguage·Claude 3.5 Sonnetmath·gpt-oss-20breasoning·Claude 3.5 Sonnetcoding·gpt-oss-20b
Hype vs Reality
Attention vs performance
Claude 3.5 Sonnet
#190 by perf·no signal
gpt-oss-20b
#97 by perf·no signal
Vendor risk
Who is behind the model
Anthropic
$965.0B·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
10 benchmarks · 2 models
Claude 3.5 Sonnetgpt-oss-20b
Chatbot Arena Elo · Overall
Claude 3.5 Sonnet leads by +56.4
Claude 3.5 Sonnet
1374.0
gpt-oss-20b
1317.6
Dtbench
gpt-oss-20b leads by +0.4
Claude 3.5 Sonnet
46.3
gpt-oss-20b
46.7
GPQA diamond
gpt-oss-20b leads by +9.0
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Claude 3.5 Sonnet
38.7
gpt-oss-20b
47.7
HELM · GPQA
gpt-oss-20b leads by +2.9
Claude 3.5 Sonnet
56.5
gpt-oss-20b
59.4
HELM · IFEval
Claude 3.5 Sonnet leads by +12.4
Claude 3.5 Sonnet
85.6
gpt-oss-20b
73.2
HELM · MMLU-Pro
Claude 3.5 Sonnet leads by +3.7
Claude 3.5 Sonnet
77.7
gpt-oss-20b
74.0
HELM · Omni-MATH
gpt-oss-20b leads by +28.9
Claude 3.5 Sonnet
27.6
gpt-oss-20b
56.5
HELM · WildBench
Claude 3.5 Sonnet leads by +5.5
Claude 3.5 Sonnet
79.2
gpt-oss-20b
73.7
OTIS Mock AIME 2024-2025
gpt-oss-20b leads by +58.8
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Claude 3.5 Sonnet
6.4
gpt-oss-20b
65.2
WeirdML
gpt-oss-20b leads by +10.0
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Claude 3.5 Sonnet
31.0
gpt-oss-20b
40.9
Full benchmark table
| Benchmark | Claude 3.5 Sonnet | gpt-oss-20b |
|---|---|---|
Chatbot Arena Elo · Overall | 1374.0 | 1317.6 |
Dtbench | 46.3 | 46.7 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 38.7 | 47.7 |
HELM · GPQA | 56.5 | 59.4 |
HELM · IFEval | 85.6 | 73.2 |
HELM · MMLU-Pro | 77.7 | 74.0 |
HELM · Omni-MATH | 27.6 | 56.5 |
HELM · WildBench | 79.2 | 73.7 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 6.4 | 65.2 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 31.0 | 40.9 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| — | — | — | — | |
| $0.02 | $0.09 | 131K tokens (~66 books) | $0.36 |