Noometry

Model comparison

Claude Sonnet 4.5 vs gpt-oss-20b

Claude Sonnet 4.5 is the stronger model overall, scoring 44.0 to 32.1 on the Noometry Index. gpt-oss-20b costs 167× less per token, which makes it the better buy when Claude Sonnet 4.5's lead doesn't matter for your workload.

Last verified . 34 shared benchmarks.

Claude Sonnet 4.5 Anthropic

44.0

Rank #76 Confirmed

gpt-oss-20b OpenAI

32.1

Rank #249 Confirmed

Summary

  • They share 34 benchmarks with published results for both. Claude Sonnet 4.5 scores higher in 8 categories and gpt-oss-20b in 1 category; 9 gaps are clear of the uncertainty.
  • The widest gap is in writing & preference, where Claude Sonnet 4.5 leads 66.7 to 35.8.
  • The biggest single-benchmark swing is Terminal-Bench: 46.5% for Claude Sonnet 4.5 and 3.4% for gpt-oss-20b.
  • gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.5.
  • Claude Sonnet 4.5 accepts more context: 1M tokens versus 131K.
  • gpt-oss-20b has downloadable open weights; the other is API-only.

Side by side

Claude Sonnet 4.5gpt-oss-20b
ProviderAnthropicOpenAI
Noometry Index44.032.1
Released2025-09-292025-08-05
WeightsProprietaryOpen
Context window1M131K
Max output64K16K
Input $ / M tokens$3$0.018
Output $ / M tokens$15$0.09
Results tracked7334

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 47.8 (#54), gpt-oss-20b: 37.1 (#185)

BenchmarkClaude Sonnet 4.5gpt-oss-20b
SciCode44.7%34.4%
WeirdML47.7%40.9%
LMArena Coding14891307
ALE-Bench796.15566.05
SWE-bench Verified71.3%—
SWE-bench Verified (bash only)71.4%—
LMArena WebDev1393—
SWE-bench Multilingual67%—
GSO14.7%—
AlgoTune1.52—

Agentic & Tool Use Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 38.1 (#30), gpt-oss-20b: 9.2 (#146)

BenchmarkClaude Sonnet 4.5gpt-oss-20b
Terminal-Bench46.5%3.4%
Berkeley Function Calling Leaderboard73.2%—
GDPval42.5%—
Remote Labor Index2.1%—
τ²-bench Airline72%—
τ²-bench Banking25.3%—
τ²-bench Retail72.4%—
τ²-bench Telecom84.9%—
Cybench60%—
DeepResearch Bench52.6%—
OSWorld62.9%—
LMArena Search1159—
METR Time Horizons67.4%—
Vending-Bench 23,839—

Reasoning Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 26.9 (#113), gpt-oss-20b: 18.8 (#251)

BenchmarkClaude Sonnet 4.5gpt-oss-20b
Kagi LLM Benchmark57.9%53.2%
CritPt1.1%1.4%
Chess Puzzles12%4%
LMArena Hard Prompts14621274
DTBench83.2%68%
LMCA38.8%14.5%
Epoch Capabilities Index146.84137.81
ARC-AGI-213.6%—
SimpleBench54.3%—
NYT Connections (extended)37.3%—
ARC-AGI-163.7%—
EnigmaEval6%—
EBR-Bench2.4%—
Mystery Game Puzzles17%—
ForecastBench61.9—

Math gpt-oss-20b leads

Claude Sonnet 4.5: 31.9 (#190), gpt-oss-20b: 37.6 (#93)

BenchmarkClaude Sonnet 4.5gpt-oss-20b
OTIS Mock AIME 2024-202577.8%65.3%
Omni-MATH55.3%56.5%
LMArena Math14481317
FrontierMath (Tiers 1-3)23.9%—
FrontierMath Tier 42.4%—
ProofBench19%—
MATH Level 597.7%—
FrontierMath (Feb 2025 set)15.2%—
FrontierMath Tier 4 (v1)4.2%—

Knowledge Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 48.4 (#70), gpt-oss-20b: 34.6 (#186)

BenchmarkClaude Sonnet 4.5gpt-oss-20b
GPQA Diamond82.3%60.8%
MMLU-Pro86.9%74%
GPQA (HELM)68.6%59.4%
LMArena Expert14831258
Humanity's Last Exam13.7%—
SimpleQA Verified30.7%—
Vectara Hallucination Rate12%—

Multimodal Not comparable

Claude Sonnet 4.5: 34.1 (#84), gpt-oss-20b: —

BenchmarkClaude Sonnet 4.5gpt-oss-20b
VPCT39.8%—
LMArena Document1450—

Multilingual Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 53.6 (#65), gpt-oss-20b: 42.4 (#192)

BenchmarkClaude Sonnet 4.5gpt-oss-20b
LMArena Non-English14251269
LMArena Chinese14591311
LMArena German14271249
LMArena Japanese13901245
LMArena Korean14031239
LMArena Russian14361279
LMArena Spanish14571267
LMArena French1459—

Instruction Following Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 75.2 (#74), gpt-oss-20b: 62.0 (#235)

BenchmarkClaude Sonnet 4.5gpt-oss-20b
IFEval85%73.2%
LMArena Instruction Following14591236

Long Context Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 45.2 (#45), gpt-oss-20b: 37.9 (#203)

BenchmarkClaude Sonnet 4.5gpt-oss-20b
LMArena Longer Query14761251

Writing & Preference Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 66.7 (#33), gpt-oss-20b: 35.8 (#260)

BenchmarkClaude Sonnet 4.5gpt-oss-20b
LMArena Text14391288
LMArena Creative Writing14421201
EQ-Bench Creative Writing1678666
WildBench85.4%73.7%
LMArena Multi-Turn14651268

Frequently asked questions

Is Claude Sonnet 4.5 better than gpt-oss-20b?

Claude Sonnet 4.5 is the stronger model overall, scoring 44.0 to 32.1 on the Noometry Index. gpt-oss-20b costs 167× less per token, which makes it the better buy when Claude Sonnet 4.5's lead doesn't matter for your workload.

Which is cheaper, Claude Sonnet 4.5 or gpt-oss-20b?

gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Claude Sonnet 4.5 lists at $3 and $15.

Is Claude Sonnet 4.5 or gpt-oss-20b better for coding?

Claude Sonnet 4.5 scores higher on coding benchmarks: 47.8 versus 37.1 in the Noometry coding category.

Which has the bigger context window?

Claude Sonnet 4.5 does, with 1M tokens against 131K.

How many benchmarks do Claude Sonnet 4.5 and gpt-oss-20b share?

34 benchmarks have published results for both models. Claude Sonnet 4.5 has 73 scored results on Noometry and gpt-oss-20b has 34.

Related comparisons

Go deeper