LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

GPT-5 Mini vs o3

Updated October 21, 2026. 86 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace GPT-5 Mini Replace o3
GPT-5 MiniOpenAI
Coding
59.8
Overall
20.6
Output $/M
$2.00
26
o3OpenAI
Coding
70.7
Overall
20.2
Output $/M
$8.00
59
vs boards won of 86 shared, 1 tied

Decision reading

Pick o3 for coding: it scores 70.7 on the Vibe Coding Index against 59.8 for GPT-5 Mini.

Pick GPT-5 Mini when output tokens cost 4.0x less, or when a solved task costs $0.08 against $0.571.

Model AGPT-5 MiniOpenAI59.8/1001 independent benchmark$0.08 per solved taskModel Bo3OpenAI70.7/1002 independent benchmarks$0.571 per solved task
Shared results
86
GPT-5 Mini only
12
o3 only
11
Like-for-like categories
8 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-5 Mini 59.8 from 1 benchmark, o3 70.7 from 2 benchmarks.
Widest category
Web design, 11.0 points apart on average over 5 shared benchmarks.
Widest single gap
SimpleQA Verified: GPT-5 Mini 21.6% against o3 49.4%, 27.8 points apart on a 0 to 100 scale.
Output price
GPT-5 Mini $2.00 against o3 $8.00 per million output tokens, 4.0x apart.
Context window
GPT-5 Mini takes 400K tokens, o3 200K.

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-5 MiniBarso3Ties
Coding43
Intelligence11
Agentic work01
Reasoning213
Math82
Knowledge391 tie
Web design41
General429
All shared boards26591 tie
CodingLike-for-likeFour of these boards feed the Vibe Coding Index43
Coding benchmarks for GPT-5 Mini and o3
BenchmarkGPT-5 Minio3Leader
Aider polyglotNot measured81.3%Only o3 measured
SWE-bench Verified59.8%58.4%GPT-5 Mini
SWE-bench MultimodalNot measured36.0%Only o3 measured
SWE-bench Multilingual39.7%Not measuredOnly GPT-5 Mini measured
LiveCodeBench (Vals AI)86.6%83.9%GPT-5 Mini
SWE-bench Verified (Vals AI)60.8%Not measuredOnly GPT-5 Mini measured
Vibe Code Bench (Vals AI)14.2%Not measuredOnly GPT-5 Mini measured
LiveCodeBench (Artificial Analysis run)83.8%80.8%GPT-5 Mini
SciCode39.0%Not measuredOnly GPT-5 Mini measured
Terminal-Bench Hard (Artificial Analysis run)33.3%37.1%o3
Terminal-Bench34.8%Not measuredOnly GPT-5 Mini measured
GSO-BenchNot measured8.8%Only o3 measured
WeirdML52.7%52.4%GPT-5 Mini
ALE-BenchNot measured933.5Only o3 measured
LMArena Text: coding14061407o3
LMArena Text: software and IT13941415o3
Neither model has a result on the other 22 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index11
Intelligence benchmarks for GPT-5 Mini and o3
BenchmarkGPT-5 Minio3Leader
AA Intelligence Index20.620.2GPT-5 Mini
Epoch Capabilities Index145.3146.7o3
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for GPT-5 Mini and o3
BenchmarkGPT-5 Minio3Leader
τ²-Bench Telecom71.1%80.7%o3
METR time horizonsNot measured65.4%Only o3 measured
GDPvalNot measured30.8%Only o3 measured
DeepResearch BenchNot measured45.2%Only o3 measured
OSWorldNot measured23.0%Only o3 measured
Vending-Bench 2-$31Not measuredOnly GPT-5 Mini measured
Neither model has a result on the other 33 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index213
Reasoning benchmarks for GPT-5 Mini and o3
BenchmarkGPT-5 Minio3Leader
GPQA Diamond80.3%84.1%o3
MMMU Pro (Vals AI)78.9%80.4%o3
GPQA Diamond (Artificial Analysis run)82.8%82.7%GPT-5 Mini
Humanity's Last Exam (Artificial Analysis run)21.5%20.1%GPT-5 Mini
GPQA Diamond (Epoch AI run)75.0%81.8%o3
ARC-AGI54.3%60.8%o3
ARC-AGI-24.4%6.5%o3
SimpleBenchNot measured53.1%Only o3 measured
Chess puzzles24.0%37.0%o3
Mystery game puzzles10.0%29.0%o3
VPCT40.2%52.0%o3
LMArena Text: hard prompts13801402o3
LMArena Text: expert13791402o3
LMArena Text: hard prompts in English13831411o3
LMArena Vision: diagrams12021220o3
LMArena Vision: homework12101234o3
Neither model has a result on the other 5 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index82
Math benchmarks for GPT-5 Mini and o3
BenchmarkGPT-5 Minio3Leader
AIME91.5%85.3%GPT-5 Mini
MATH 500 (Vals AI)94.8%94.6%GPT-5 Mini
MGSM (Vals AI)92.6%91.8%GPT-5 Mini
AA Math Index90.788.3GPT-5 Mini
AIME 2025 (Artificial Analysis run)90.7%88.3%GPT-5 Mini
OTIS Mock AIME86.7%84.4%GPT-5 Mini
FrontierMath, tiers 1 to 346.7%33.3%GPT-5 Mini
FrontierMath, tier 412.2%Not measuredOnly GPT-5 Mini measured
ProofBench9.0%Not measuredOnly GPT-5 Mini measured
MATH level 597.8%97.8%GPT-5 Mini
LMArena Text: math13781426o3
LMArena Text: mathematical work13801423o3
Neither model has a result on the other 1 board here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index39
Knowledge benchmarks for GPT-5 Mini and o3
BenchmarkGPT-5 Minio3Leader
MMLU Pro82.2%85.6%o3
LegalBench (Vals AI)81.8%83.8%o3
CaseLaw (Vals AI)68.5%Not measuredOnly GPT-5 Mini measured
CorpFin (Vals AI)60.2%59.7%GPT-5 Mini
TaxEval (Vals AI)75.2%74.6%GPT-5 Mini
MortgageTax (Vals AI)66.9%65.7%GPT-5 Mini
MedQA (Vals AI)96.1%96.1%Tie
MedCode (Vals AI)43.0%47.3%o3
MMLU Pro (Artificial Analysis run)83.7%85.3%o3
SimpleQA Verified21.6%49.4%o3
GeoBenchNot measured74.0%Only o3 measured
LMArena Text: science13711429o3
LMArena Text: medicine and healthcare13801448o3
LMArena Text: legal and government13751428o3
LMArena Text: business and finance13581393o3
Neither model has a result on the other 2 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index41
Web design benchmarks for GPT-5 Mini and o3
BenchmarkGPT-5 Minio3Leader
Design Arena: websites11251036GPT-5 Mini
Design Arena: UI components11061016GPT-5 Mini
Design Arena: game dev11301034GPT-5 Mini
Design Arena: data viz11311200o3
Design Arena: 3D1037Not measuredOnly GPT-5 Mini measured
Design Arena: all code categories11151021GPT-5 Mini
Design Arena: SVG1091Not measuredOnly GPT-5 Mini measured
Design Arena: ASCII art1124Not measuredOnly GPT-5 Mini measured
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index429
General benchmarks for GPT-5 Mini and o3
BenchmarkGPT-5 Minio3Leader
SAGE (Vals AI)43.0%41.8%GPT-5 Mini
MedScribe (Vals AI)80.6%76.7%GPT-5 Mini
IFBench75.4%71.4%GPT-5 Mini
Long context reasoning (AA-LCR)72.3%74.7%o3
DTBench80.5%84.8%o3
LMCA34.2%39.7%o3
Fiction.LiveBench69.4%88.9%o3
Creative writing (Lech Mazur)83.1%83.9%o3
CL-benchNot measured17.8%Only o3 measured
LMArena Text13731410o3
LMArena Text: instruction following13571368o3
LMArena Text: creative writing13251359o3
LMArena Vision12021214o3
LMArena Text: English13791417o3
LMArena Text: non-English13631401o3
LMArena Text: Chinese13861437o3
LMArena Text: German13661420o3
LMArena Text: Japanese13411403o3
LMArena Text: Korean13081370o3
LMArena Text: French13861430o3
LMArena Text: Spanish13541395o3
LMArena Text: Russian13621406o3
LMArena Text: Polish13781438o3
LMArena Text: multi-turn13631405o3
LMArena Text: longer queries13551372o3
LMArena Text: writing and language13441372o3
LMArena Text: entertainment and media13401373o3
LMArena Vision: English12161221o3
LMArena Vision: Chinese12141237o3
LMArena Vision: reading text11991218o3
LMArena Vision: humor12051222o3
LMArena Vision: captioning12171210GPT-5 Mini
LMArena Vision: recognizing things12261234o3
LMArena Vision: creative writing11811201o3
Neither model has a result on the other 12 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-5 Mini and o3
BenchmarkGPT-5 Minio3Leader
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-5 Mini$0.0013Fits in one request
o3$0.0060Fits in one request

GPT-5 Mini costs less

Repository review

50K input + 3K output tokens

GPT-5 Mini$0.018Fits in one request
o3$0.124Fits in one request

GPT-5 Mini costs less

Agent loop

200K input + 10K output tokens

GPT-5 Mini$0.070Fits in one request
o3$0.480Exceeds the context window

GPT-5 Mini 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-5 Mini and o3
SpecificationOpenAI GPT-5 MiniOpenAI o3
MakerOpenAIOpenAI
WeightsNo public weights listedNo public weights listed
ReleasedAugust 7, 2025April 16, 2025
Context window400K200K
Input $/Mtok$0.25$2.00
Output $/Mtok$2.00$8.00
Coding index59.870.7
Independent measurements behind it12
Cost per solved task$0.08$0.571
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-5 Mini better than o3 for coding?

Pick o3 for coding: it scores 70.7 on the Vibe Coding Index against 59.8 for GPT-5 Mini. Pick GPT-5 Mini when output tokens cost 4.0x less, or when a solved task costs $0.08 against $0.571.

Which is cheaper, GPT-5 Mini or o3?

GPT-5 Mini costs $0.25 per million input tokens and $2.00 per million output tokens; o3 costs $2.00 and $8.00. Per solved SWE-bench Verified task, GPT-5 Mini costs $0.08 and o3 $0.571.

Which has the bigger context window, GPT-5 Mini or o3?

GPT-5 Mini, with 400K tokens against 200K.

How many benchmarks is the GPT-5 Mini vs o3 comparison based on?

86 benchmarks have a result for both models as of October 21, 2026. GPT-5 Mini's coding index rests on 1 independent benchmark and o3's on 2. Where only one model has a result, the page says so instead of naming a winner.

Compare next