GPT-5 vs GPT-5.1

Updated October 21, 2026. 94 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace GPT-5 Replace GPT-5.1

GPT-5OpenAI

Coding: 72.2
Overall: 23.0
Output $/M: $10.00

GPT-5.1OpenAI

Coding: 53.3
Overall: 24.7
Output $/M: $10.00

vs boards won of 94 shared, 1 tied

Decision reading

Pick GPT-5 for coding: it scores 72.2 on the Vibe Coding Index against 53.3 for GPT-5.1.

Pick GPT-5.1 when what matters is AA Coding Index or LiveCodeBench (Vals AI), where it leads.

Model AGPT-5OpenAI72.2/1003 independent benchmarks$0.39 per solved task Model BGPT-5.1OpenAI53.3/1002 independent benchmarks$1.02 per solved task

94156

Shared results: 94
GPT-5 only: 15
GPT-5.1 only: 6
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.

Coding
Vibe Coding Index, four public coding benchmarks
GPT-5
GPT-5 scores 72.2 against 53.3.
3 and 2 benchmarks behind the scores
Intelligence
AA Intelligence Index
GPT-5.1
GPT-5.1 leads, 24.7 against 23.0.
Same board, both measured
Reasoning
GPQA Diamond
GPT-5.1
GPT-5.1 leads, 86.6% against 85.6%.
Same board, both measured
Math
AIME
GPT-5
GPT-5 leads, 93.4% against 93.3%.
Same board, both measured
Knowledge
MMLU Pro
GPT-5
GPT-5 leads, 86.5% against 86.4%.
Same board, both measured
Web design
Design Arena: websites
GPT-5.1
GPT-5.1 leads, 1187 against 1185.
Same board, both measured
Long documents
Listed context window
No call
Both take 400K tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
No call
GPT-5 $0.0063, GPT-5.1 $0.0063 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
GPT-5
GPT-5 $0.39, GPT-5.1 $1.02.
Published attempt cost and pass rate

At a glance

Coding index: GPT-5 72.2 from 3 benchmarks, GPT-5.1 53.3 from 2 benchmarks.
Widest category: Agentic work, 5.7 points apart on average over 2 shared benchmarks.
Widest single gap: LMArena Text: Chinese: GPT-5 1422 against GPT-5.1 1495, 12.2 points apart on a 0 to 100 scale.
Output price: Both cost $10.00 per million output tokens.
Context window: Both take 400K tokens.

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
Category	GPT-5	GPT-5.1	Ties
Coding	4	10
Intelligence	1	1
Agentic work	2	0
Reasoning	4	11	1 tie
Math	4	3
Knowledge	5	9
Web design	5	3
General	4	27
All shared boards	29	64	1 tie

CodingLike-for-likeFour of these boards feed the Vibe Coding Index410

Coding benchmarks for GPT-5 and GPT-5.1
Benchmark	GPT-5	GPT-5.1	Leader
Aider polyglot	88.0%	Not measured	Only GPT-5 measured
SWE-bench Verified	74.4%	66.0%	GPT-5
LMArena WebDev	1418	1396	GPT-5
AA Coding Index	37.8	49.4	GPT-5.1
LiveCodeBench (Vals AI)	85.9%	86.5%	GPT-5.1
SWE-bench Verified (Vals AI)	69.0%	69.8%	GPT-5.1
Vibe Code Bench (Vals AI)	20.1%	24.6%	GPT-5.1
LiveCodeBench (Artificial Analysis run)	84.6%	86.8%	GPT-5.1
Terminal-Bench Hard (Artificial Analysis run)	37.9%	45.5%	GPT-5.1
Terminal-Bench	49.6%	47.6%	GPT-5
GSO-Bench	6.9%	13.7%	GPT-5.1
WeirdML	60.7%	60.8%	GPT-5.1
ALE-Bench	807.6	Not measured	Only GPT-5 measured
LMArena Text: coding	1436	1453	GPT-5.1
LMArena Text: software and IT	1420	1454	GPT-5.1
LMArena WebDev: React	Not measured	1307	Only GPT-5.1 measured
LMArena WebDev: HTML	1427	1405	GPT-5
LMArena WebDev: from an image	Not measured	1421	Only GPT-5.1 measured
Neither model has a result on the other 20 boards here

IntelligenceLike-for-likePublic boards, shown as published, kept out of the index11

Intelligence benchmarks for GPT-5 and GPT-5.1
Benchmark	GPT-5	GPT-5.1	Leader
AA Intelligence Index	23.0	24.7	GPT-5.1
Epoch Capabilities Index	150.0	149.8	GPT-5

Agentic workLike-for-likePublic boards, shown as published, kept out of the index20

Agentic work benchmarks for GPT-5 and GPT-5.1
Benchmark	GPT-5	GPT-5.1	Leader
τ²-Bench Telecom	86.5%	81.9%	GPT-5
METR time horizons	69.6%	Not measured	Only GPT-5 measured
Remote Labor Index	1.7%	Not measured	Only GPT-5 measured
GDPval	34.8%	Not measured	Only GPT-5 measured
BALROG	32.8%	Not measured	Only GPT-5 measured
DeepResearch Bench	49.6%	42.8%	GPT-5
Vending-Bench 2	Not measured	$1,473	Only GPT-5.1 measured
Neither model has a result on the other 32 boards here

ReasoningLike-for-likePublic boards, shown as published, kept out of the index411

Reasoning benchmarks for GPT-5 and GPT-5.1
Benchmark	GPT-5	GPT-5.1	Leader
GPQA Diamond	85.6%	86.6%	GPT-5.1
MMMU Pro (Vals AI)	81.5%	83.2%	GPT-5.1
GPQA Diamond (Artificial Analysis run)	85.4%	87.3%	GPT-5.1
Humanity's Last Exam (Artificial Analysis run)	28.5%	28.5%	Tie
GPQA Diamond (Epoch AI run)	86.2%	87.6%	GPT-5.1
ARC-AGI	65.7%	72.8%	GPT-5.1
ARC-AGI-2	9.9%	17.6%	GPT-5.1
SimpleBench	56.7%	53.2%	GPT-5
Chess puzzles	36.0%	42.0%	GPT-5.1
Mystery game puzzles	23.0%	19.0%	GPT-5
EBR-bench	12.7%	Not measured	Only GPT-5 measured
VPCT	66.0%	58.7%	GPT-5
LMArena Text: hard prompts	1415	1457	GPT-5.1
LMArena Text: expert	1419	1470	GPT-5.1
LMArena Text: hard prompts in English	1419	1462	GPT-5.1
LMArena Vision: diagrams	1252	1264	GPT-5.1
LMArena Vision: homework	1278	1272	GPT-5
Neither model has a result on the other 4 boards here

MathLike-for-likePublic boards, shown as published, kept out of the index43

Math benchmarks for GPT-5 and GPT-5.1
Benchmark	GPT-5	GPT-5.1	Leader
AIME	93.4%	93.3%	GPT-5
MATH 500 (Vals AI)	96.0%	Not measured	Only GPT-5 measured
MGSM (Vals AI)	92.8%	93.0%	GPT-5.1
AA Math Index	94.3	94.0	GPT-5
AIME 2025 (Artificial Analysis run)	94.3%	94.0%	GPT-5
OTIS Mock AIME	91.4%	88.6%	GPT-5
FrontierMath, tiers 1 to 3	55.4%	Not measured	Only GPT-5 measured
FrontierMath, tier 4	22.0%	Not measured	Only GPT-5 measured
ProofBench	18.0%	Not measured	Only GPT-5 measured
MATH level 5	98.1%	Not measured	Only GPT-5 measured
LMArena Text: math	1407	1446	GPT-5.1
LMArena Text: mathematical work	1405	1449	GPT-5.1
Neither model has a result on the other 1 board here

KnowledgeLike-for-likePublic boards, shown as published, kept out of the index59

Knowledge benchmarks for GPT-5 and GPT-5.1
Benchmark	GPT-5	GPT-5.1	Leader
MMLU Pro	86.5%	86.4%	GPT-5
LegalBench (Vals AI)	86.0%	85.7%	GPT-5
CaseLaw (Vals AI)	66.5%	73.4%	GPT-5.1
CorpFin (Vals AI)	61.1%	63.8%	GPT-5.1
TaxEval (Vals AI)	73.4%	74.9%	GPT-5.1
MortgageTax (Vals AI)	65.5%	61.4%	GPT-5
MedQA (Vals AI)	96.3%	96.4%	GPT-5.1
MedCode (Vals AI)	49.6%	52.7%	GPT-5.1
MMLU Pro (Artificial Analysis run)	87.1%	87.0%	GPT-5
SimpleQA Verified	50.1%	48.0%	GPT-5
GeoBench	81.0%	Not measured	Only GPT-5 measured
LMArena Text: science	1416	1459	GPT-5.1
LMArena Text: medicine and healthcare	1413	1456	GPT-5.1
LMArena Text: legal and government	1431	1459	GPT-5.1
LMArena Text: business and finance	1400	1431	GPT-5.1
Neither model has a result on the other 2 boards here

Web designLike-for-likePublic boards, shown as published, kept out of the index53

Web design benchmarks for GPT-5 and GPT-5.1
Benchmark	GPT-5	GPT-5.1	Leader
Design Arena: websites	1185	1187	GPT-5.1
Design Arena: UI components	1179	1166	GPT-5
Design Arena: game dev	1185	1179	GPT-5
Design Arena: data viz	1221	1204	GPT-5
Design Arena: 3D	1056	1061	GPT-5.1
Design Arena: all code categories	1168	1171	GPT-5.1
Design Arena: SVG	1188	1149	GPT-5
Design Arena: ASCII art	1140	1118	GPT-5
Neither model has a result on the other 1 board here

GeneralLike-for-likePublic boards, shown as published, kept out of the index427

General benchmarks for GPT-5 and GPT-5.1
Benchmark	GPT-5	GPT-5.1	Leader
SAGE (Vals AI)	43.7%	43.2%	GPT-5
MedScribe (Vals AI)	83.7%	88.1%	GPT-5.1
IFBench	73.1%	72.9%	GPT-5
Long context reasoning (AA-LCR)	78.2%	80.0%	GPT-5.1
DTBench	90.7%	90.1%	GPT-5
LMCA	40.0%	43.9%	GPT-5.1
Fiction.LiveBench	97.2%	Not measured	Only GPT-5 measured
Creative writing (Lech Mazur)	86.0%	Not measured	Only GPT-5 measured
CL-bench	Not measured	23.7%	Only GPT-5.1 measured
CL-bench Life	Not measured	17.3%	Only GPT-5.1 measured
LMArena Text	1406	1443	GPT-5.1
LMArena Text: instruction following	1388	1443	GPT-5.1
LMArena Text: creative writing	1365	1427	GPT-5.1
LMArena Vision	1232	1250	GPT-5.1
LMArena Text: English	1409	1451	GPT-5.1
LMArena Text: non-English	1397	1431	GPT-5.1
LMArena Text: Chinese	1422	1495	GPT-5.1
LMArena Text: German	1416	1438	GPT-5.1
LMArena Text: Japanese	1409	1453	GPT-5.1
LMArena Text: Korean	1360	1401	GPT-5.1
LMArena Text: French	1409	1450	GPT-5.1
LMArena Text: Spanish	1399	1433	GPT-5.1
LMArena Text: Russian	1407	1435	GPT-5.1
LMArena Text: Polish	1431	1440	GPT-5.1
LMArena Text: multi-turn	1426	1450	GPT-5.1
LMArena Text: longer queries	1399	1448	GPT-5.1
LMArena Text: writing and language	1384	1428	GPT-5.1
LMArena Text: entertainment and media	1379	1415	GPT-5.1
LMArena Vision: English	1231	1252	GPT-5.1
LMArena Vision: Chinese	1267	1270	GPT-5.1
LMArena Vision: reading text	1246	1255	GPT-5.1
LMArena Vision: humor	1219	1263	GPT-5.1
LMArena Vision: captioning	1212	1252	GPT-5.1
LMArena Vision: recognizing things	1263	1232	GPT-5
LMArena Vision: creative writing	1210	1266	GPT-5.1
LMArena Document	Not measured	1403	Only GPT-5.1 measured
Neither model has a result on the other 10 boards here

Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared

Image generation benchmarks for GPT-5 and GPT-5.1
Benchmark	GPT-5	GPT-5.1	Leader
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$0.0063Fits in one request

GPT-5.1$0.0063Fits in one request

Same cost

Repository review

50K input + 3K output tokens

GPT-5$0.092Fits in one request

GPT-5.1$0.092Fits in one request

Same cost

Agent loop

200K input + 10K output tokens

GPT-5$0.350Fits in one request

GPT-5.1$0.350Fits in one request

Same cost

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 and GPT-5.1
Specification	OpenAI GPT-5	OpenAI GPT-5.1
Maker	OpenAI	OpenAI
Weights	No public weights listed	No public weights listed
Released	August 7, 2025	November 13, 2025
Context window	400K	400K
Input $/Mtok	$1.25	$1.25
Output $/Mtok	$10.00	$10.00
Coding index	72.2	53.3
Independent measurements behind it	3	2
Cost per solved task	$0.39	$1.02
Source	OpenRouter listing	OpenRouter listing

Questions

Is GPT-5 better than GPT-5.1 for coding?

Pick GPT-5 for coding: it scores 72.2 on the Vibe Coding Index against 53.3 for GPT-5.1. Pick GPT-5.1 when what matters is AA Coding Index or LiveCodeBench (Vals AI), where it leads.

Which is cheaper, GPT-5 or GPT-5.1?

GPT-5 costs $1.25 per million input tokens and $10.00 per million output tokens; GPT-5.1 costs $1.25 and $10.00. Per solved SWE-bench Verified task, GPT-5 costs $0.39 and GPT-5.1 $1.02.

Which has the bigger context window, GPT-5 or GPT-5.1?

Neither: both take 400K tokens.

How many benchmarks is the GPT-5 vs GPT-5.1 comparison based on?

94 benchmarks have a result for both models as of October 21, 2026. GPT-5's coding index rests on 3 independent benchmarks and GPT-5.1's on 2. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/gpt-5-vs-gpt-5-1" title="GPT-5 vs GPT-5.1, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/gpt-5-vs-gpt-5-1">LLMBenchmark.ai, GPT-5 vs GPT-5.1</a></p>

Decision reading

Which one for your work

Coding

Intelligence

Reasoning

Math

Knowledge

Web design

Long documents

Everyday cost

Cost per solved task

At a glance

Benchmark by benchmark

What each workload costs

Chat turn

Repository review

Agent loop

Specification differences

Questions

Compare next