LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

GPT-4 vs gpt-oss-120b

Updated October 21, 2026. 37 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace GPT-4 Replace gpt-oss-120b
GPT-4OpenAI
Coding
22.4
Overall
6.7
Output $/M
$60.00
0
gpt-oss-120bOpenAI
Coding
34.5
Overall
11.6
Output $/M
$0.60
37
vs boards won of 37 shared

Decision reading

Pick gpt-oss-120b for coding: it scores 34.5 on the Vibe Coding Index against 22.4 for GPT-4.

GPT-4 is neither cheaper nor larger in context, and leads no benchmark both models were measured on.

Model AGPT-4OpenAI22.4/1001 independent benchmarkNo cost per solved taskModel Bgpt-oss-120bOpenAI34.5/1002 independent benchmarks$0.51 per solved task
Shared results
37
GPT-4 only
8
gpt-oss-120b only
41
Like-for-like categories
6 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
GPT-4 22.4 from 1 benchmark, gpt-oss-120b 34.5 from 2 benchmarks.
Widest category
Reasoning, 32.9 points apart on average over 4 shared benchmarks.
Widest single gap
GPQA Diamond (Artificial Analysis run): GPT-4 34.9% against gpt-oss-120b 78.2%, 43.3 points apart on a 0 to 100 scale.
Output price
GPT-4 $60.00 against gpt-oss-120b $0.60 per million output tokens, 100.0x apart.
Context window
GPT-4 takes 8K tokens, gpt-oss-120b 131K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryGPT-4Barsgpt-oss-120bTies
Coding04
Intelligence02
Reasoning04
Math02
Knowledge05
General020
All shared boards037
CodingLike-for-likeFour of these boards feed the Vibe Coding Index04
Coding benchmarks for GPT-4 and gpt-oss-120b
BenchmarkGPT-4gpt-oss-120bLeader
Aider polyglotNot measured41.8%Only gpt-oss-120b measured
SWE-bench Verified22.4%26.0%gpt-oss-120b
SWE-bench Lite39.7%Not measuredOnly GPT-4 measured
SWE-bench (full test set)12.5%Not measuredOnly GPT-4 measured
AA Coding Index13.130.4gpt-oss-120b
LiveCodeBench (Vals AI)Not measured83.2%Only gpt-oss-120b measured
SWE-bench Verified (Vals AI)Not measured33.6%Only gpt-oss-120b measured
LiveCodeBench (Artificial Analysis run)Not measured87.8%Only gpt-oss-120b measured
SciCodeNot measured34.0%Only gpt-oss-120b measured
Terminal-Bench Hard (Artificial Analysis run)Not measured23.5%Only gpt-oss-120b measured
Terminal-BenchNot measured18.7%Only gpt-oss-120b measured
WeirdMLNot measured48.2%Only gpt-oss-120b measured
ALE-BenchNot measured575.6Only gpt-oss-120b measured
LMArena Text: coding12091379gpt-oss-120b
LMArena Text: software and IT11951387gpt-oss-120b
Neither model has a result on the other 23 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for GPT-4 and gpt-oss-120b
BenchmarkGPT-4gpt-oss-120bLeader
AA Intelligence Index6.711.6gpt-oss-120b
Epoch Capabilities Index125.7139.8gpt-oss-120b
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index5 boards, none shared
Agentic work benchmarks for GPT-4 and gpt-oss-120b
BenchmarkGPT-4gpt-oss-120bLeader
AA Agentic IndexNot measured3.7Only gpt-oss-120b measured
τ²-Bench TelecomNot measured65.8%Only gpt-oss-120b measured
APEX-AgentsNot measured4.4%Only gpt-oss-120b measured
METR time horizonsNot measured56.6%Only gpt-oss-120b measured
Vending-Bench 2Not measured-$22Only gpt-oss-120b measured
Neither model has a result on the other 34 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index04
Reasoning benchmarks for GPT-4 and gpt-oss-120b
BenchmarkGPT-4gpt-oss-120bLeader
GPQA DiamondNot measured78.5%Only gpt-oss-120b measured
GPQA Diamond (Artificial Analysis run)34.9%78.2%gpt-oss-120b
Humanity's Last Exam (Artificial Analysis run)Not measured19.6%Only gpt-oss-120b measured
GPQA Diamond (Epoch AI run)Not measured75.8%Only gpt-oss-120b measured
SimpleBenchNot measured22.1%Only gpt-oss-120b measured
Mystery game puzzlesNot measured2.0%Only gpt-oss-120b measured
BIG-Bench Hard75.1%Not measuredOnly GPT-4 measured
LMArena Text: hard prompts12001364gpt-oss-120b
LMArena Text: expert11491356gpt-oss-120b
LMArena Text: hard prompts in English12091368gpt-oss-120b
Neither model has a result on the other 11 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index02
Math benchmarks for GPT-4 and gpt-oss-120b
BenchmarkGPT-4gpt-oss-120bLeader
AIMENot measured92.6%Only gpt-oss-120b measured
MATH 500 (Vals AI)Not measured94.8%Only gpt-oss-120b measured
MGSM (Vals AI)Not measured92.0%Only gpt-oss-120b measured
AA Math IndexNot measured93.4Only gpt-oss-120b measured
AIME 2025 (Artificial Analysis run)Not measured93.4%Only gpt-oss-120b measured
OTIS Mock AIMENot measured88.9%Only gpt-oss-120b measured
GSM8K92.0%Not measuredOnly GPT-4 measured
LMArena Text: math12301389gpt-oss-120b
LMArena Text: mathematical work12281386gpt-oss-120b
Neither model has a result on the other 4 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index05
Knowledge benchmarks for GPT-4 and gpt-oss-120b
BenchmarkGPT-4gpt-oss-120bLeader
MMLU ProNot measured79.2%Only gpt-oss-120b measured
LegalBench (Vals AI)Not measured75.9%Only gpt-oss-120b measured
CaseLaw (Vals AI)Not measured48.8%Only gpt-oss-120b measured
CorpFin (Vals AI)Not measured58.2%Only gpt-oss-120b measured
TaxEval (Vals AI)Not measured71.6%Only gpt-oss-120b measured
MedQA (Vals AI)Not measured91.4%Only gpt-oss-120b measured
MMLU Pro (Artificial Analysis run)56.2%80.8%gpt-oss-120b
MMLU86.4%Not measuredOnly GPT-4 measured
AI2 Reasoning Challenge96.3%Not measuredOnly GPT-4 measured
LMArena Text: science11771382gpt-oss-120b
LMArena Text: medicine and healthcare11451388gpt-oss-120b
LMArena Text: legal and government11911355gpt-oss-120b
LMArena Text: business and finance11451359gpt-oss-120b
Neither model has a result on the other 4 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
Web design benchmarks for GPT-4 and gpt-oss-120b
BenchmarkGPT-4gpt-oss-120bLeader
Design Arena: websitesNot measured968Only gpt-oss-120b measured
Design Arena: UI componentsNot measured927Only gpt-oss-120b measured
Design Arena: game devNot measured992Only gpt-oss-120b measured
Design Arena: data vizNot measured987Only gpt-oss-120b measured
Design Arena: 3DNot measured901Only gpt-oss-120b measured
Design Arena: all code categoriesNot measured963Only gpt-oss-120b measured
Neither model has a result on the other 3 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index020
General benchmarks for GPT-4 and gpt-oss-120b
BenchmarkGPT-4gpt-oss-120bLeader
IFBench33.2%69.0%gpt-oss-120b
Long context reasoning (AA-LCR)Not measured52.0%Only gpt-oss-120b measured
DTBench62.7%76.3%gpt-oss-120b
LMCA17.1%22.1%gpt-oss-120b
Fiction.LiveBenchNot measured44.4%Only gpt-oss-120b measured
Creative writing (Lech Mazur)Not measured77.1%Only gpt-oss-120b measured
Surface Evolver BenchNot measured25.0%Only gpt-oss-120b measured
HellaSwag95.3%Not measuredOnly GPT-4 measured
WinoGrande87.5%Not measuredOnly GPT-4 measured
LMArena Text12061366gpt-oss-120b
LMArena Text: instruction following12011318gpt-oss-120b
LMArena Text: creative writing11911274gpt-oss-120b
LMArena Text: English12201375gpt-oss-120b
LMArena Text: non-English11941351gpt-oss-120b
LMArena Text: Chinese11841385gpt-oss-120b
LMArena Text: German11981353gpt-oss-120b
LMArena Text: Japanese11371331gpt-oss-120b
LMArena Text: Korean10871282gpt-oss-120b
LMArena Text: French12201370gpt-oss-120b
LMArena Text: Spanish11971389gpt-oss-120b
LMArena Text: Russian11951343gpt-oss-120b
LMArena Text: Polish12651353gpt-oss-120b
LMArena Text: multi-turn12061340gpt-oss-120b
LMArena Text: longer queries11901319gpt-oss-120b
LMArena Text: writing and language12191306gpt-oss-120b
LMArena Text: entertainment and media11841292gpt-oss-120b
Neither model has a result on the other 20 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-4 and gpt-oss-120b
BenchmarkGPT-4gpt-oss-120bLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

GPT-4$0.060Fits in one request
gpt-oss-120b$0.0004Fits in one request

gpt-oss-120b costs less

Repository review

50K input + 3K output tokens

GPT-4$1.68Exceeds the context window
gpt-oss-120b$0.0093Fits in one request

gpt-oss-120b costs less

Agent loop

200K input + 10K output tokens

GPT-4$6.60Exceeds the context window
gpt-oss-120b$0.036Exceeds the context window

gpt-oss-120b costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of GPT-4 and gpt-oss-120b
SpecificationOpenAI GPT-4OpenAI gpt-oss-120b
MakerOpenAIOpenAI
WeightsNo public weights listedOpen weights
ReleasedMay 28, 2023August 5, 2025
Context window8K131K
Input $/Mtok$30.00$0.15
Output $/Mtok$60.00$0.60
Coding index22.434.5
Independent measurements behind it12
Cost per solved taskNot computable$0.51
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-4 better than gpt-oss-120b for coding?

Pick gpt-oss-120b for coding: it scores 34.5 on the Vibe Coding Index against 22.4 for GPT-4. GPT-4 is neither cheaper nor larger in context, and leads no benchmark both models were measured on.

Which is cheaper, GPT-4 or gpt-oss-120b?

GPT-4 costs $30.00 per million input tokens and $60.00 per million output tokens; gpt-oss-120b costs $0.15 and $0.60. Cost per solved task is known for gpt-oss-120b only.

Which has the bigger context window, GPT-4 or gpt-oss-120b?

gpt-oss-120b, with 131K tokens against 8K.

How many benchmarks is the GPT-4 vs gpt-oss-120b comparison based on?

37 benchmarks have a result for both models as of October 21, 2026. GPT-4's coding index rests on 1 independent benchmark and gpt-oss-120b's on 2. Where only one model has a result, the page says so instead of naming a winner.

Compare next