LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

R1 0528 vs GPT-5.2

Updated October 21, 2026. 55 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace R1 0528 Replace GPT-5.2
R1 0528DeepSeek
Coding
71.4
Overall
13.1
Output $/M
$2.15
12
GPT-5.2OpenAI
Coding
58.9
Overall
30.4
Output $/M
$14.00
43
vs boards won of 55 shared

Decision reading

Pick R1 0528 for coding: it scores 71.4 on the Vibe Coding Index against 58.9 for GPT-5.2, though on 1 benchmark against 2.

Pick GPT-5.2 when the context window is larger, 400K against 164K tokens, or when what matters is LiveCodeBench (Artificial Analysis run) or Terminal-Bench Hard (Artificial Analysis run), where it leads.

Model AR1 0528DeepSeek71.4/1001 independent benchmarkNo cost per solved taskModel BGPT-5.2OpenAI58.9/1002 independent benchmarks$1.91 per solved task
Shared results
55
R1 0528 only
4
GPT-5.2 only
56
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
R1 0528 71.4 from 1 benchmark, GPT-5.2 58.9 from 2 benchmarks.
Widest category
Agentic work, 25.3 points apart on average over 3 shared benchmarks.
Widest single gap
ARC-AGI: R1 0528 21.2% against GPT-5.2 86.2%, 65.0 points apart on a 0 to 100 scale.
Output price
R1 0528 $2.15 against GPT-5.2 $14.00 per million output tokens, 6.5x apart.
Context window
R1 0528 takes 164K tokens, GPT-5.2 400K.

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
CategoryR1 0528BarsGPT-5.2Ties
Coding15
Intelligence01
Agentic work03
Reasoning09
Math05
Knowledge23
Web design16
General811
All shared boards1243
CodingLike-for-likeFour of these boards feed the Vibe Coding Index15
Coding benchmarks for R1 0528 and GPT-5.2
BenchmarkR1 0528GPT-5.2Leader
Aider polyglot71.4%Not measuredOnly R1 0528 measured
SWE-bench VerifiedNot measured72.8%Only GPT-5.2 measured
SWE-bench MultilingualNot measured66.7%Only GPT-5.2 measured
LMArena WebDevNot measured1417Only GPT-5.2 measured
LiveCodeBench (Vals AI)Not measured85.4%Only GPT-5.2 measured
SWE-bench Verified (Vals AI)Not measured75.8%Only GPT-5.2 measured
Vibe Code Bench (Vals AI)Not measured53.5%Only GPT-5.2 measured
LiveCodeBench (Artificial Analysis run)77.0%89.4%GPT-5.2
Terminal-Bench Hard (Artificial Analysis run)15.9%47.0%GPT-5.2
Terminal-BenchNot measured64.9%Only GPT-5.2 measured
GSO-BenchNot measured27.5%Only GPT-5.2 measured
WeirdML41.6%72.2%GPT-5.2
ALE-Bench804.11293.5GPT-5.2
LMArena Text: coding14271442GPT-5.2
LMArena Text: software and IT14351434R1 0528
LMArena WebDev: HTMLNot measured1427Only GPT-5.2 measured
Neither model has a result on the other 22 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for R1 0528 and GPT-5.2
BenchmarkR1 0528GPT-5.2Leader
AA Intelligence Index13.130.4GPT-5.2
Epoch Capabilities IndexNot measured153.7Only GPT-5.2 measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index03
Agentic work benchmarks for R1 0528 and GPT-5.2
BenchmarkR1 0528GPT-5.2Leader
Design Arena: agents building web appsNot measured1086Only GPT-5.2 measured
Design Arena: agents building full-stack appsNot measured1043Only GPT-5.2 measured
Design Arena: agents building mobile appsNot measured1094Only GPT-5.2 measured
Design Arena: agents building Android appsNot measured1064Only GPT-5.2 measured
Design Arena: agents building Godot gamesNot measured1142Only GPT-5.2 measured
τ²-Bench Telecom36.5%84.8%GPT-5.2
METR time horizons53.8%75.3%GPT-5.2
Remote Labor IndexNot measured2.5%Only GPT-5.2 measured
GDPvalNot measured49.7%Only GPT-5.2 measured
DeepResearch Bench35.1%41.1%GPT-5.2
Vending-Bench 2Not measured$3,591Only GPT-5.2 measured
Neither model has a result on the other 28 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index09
Reasoning benchmarks for R1 0528 and GPT-5.2
BenchmarkR1 0528GPT-5.2Leader
GPQA DiamondNot measured91.7%Only GPT-5.2 measured
MMMU Pro (Vals AI)Not measured86.7%Only GPT-5.2 measured
GPQA Diamond (Artificial Analysis run)81.3%90.3%GPT-5.2
Humanity's Last Exam (Artificial Analysis run)15.8%37.7%GPT-5.2
GPQA Diamond (Epoch AI run)76.3%91.4%GPT-5.2
ARC-AGI21.2%86.2%GPT-5.2
ARC-AGI-21.1%52.9%GPT-5.2
SimpleBench40.8%45.8%GPT-5.2
Chess puzzlesNot measured48.0%Only GPT-5.2 measured
Mystery game puzzlesNot measured23.0%Only GPT-5.2 measured
EBR-benchNot measured23.0%Only GPT-5.2 measured
VPCTNot measured84.0%Only GPT-5.2 measured
LMArena Text: hard prompts14151428GPT-5.2
LMArena Text: expert13941445GPT-5.2
LMArena Text: hard prompts in English14221431GPT-5.2
LMArena Vision: diagramsNot measured1265Only GPT-5.2 measured
LMArena Vision: homeworkNot measured1276Only GPT-5.2 measured
Neither model has a result on the other 4 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index05
Math benchmarks for R1 0528 and GPT-5.2
BenchmarkR1 0528GPT-5.2Leader
AIMENot measured96.9%Only GPT-5.2 measured
MGSM (Vals AI)Not measured94.0%Only GPT-5.2 measured
AA Math Index76.099.0GPT-5.2
AIME 2025 (Artificial Analysis run)76.0%99.0%GPT-5.2
OTIS Mock AIME66.4%96.1%GPT-5.2
FrontierMath, tiers 1 to 3Not measured67.4%Only GPT-5.2 measured
FrontierMath, tier 4Not measured31.7%Only GPT-5.2 measured
ProofBenchNot measured15.0%Only GPT-5.2 measured
MATH level 596.6%Not measuredOnly R1 0528 measured
LMArena Text: math14001440GPT-5.2
LMArena Text: mathematical work14151442GPT-5.2
Neither model has a result on the other 2 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index23
Knowledge benchmarks for R1 0528 and GPT-5.2
BenchmarkR1 0528GPT-5.2Leader
MMLU ProNot measured86.2%Only GPT-5.2 measured
LegalBench (Vals AI)Not measured82.8%Only GPT-5.2 measured
CaseLaw (Vals AI)Not measured66.0%Only GPT-5.2 measured
CorpFin (Vals AI)Not measured65.9%Only GPT-5.2 measured
TaxEval (Vals AI)Not measured75.8%Only GPT-5.2 measured
MortgageTax (Vals AI)Not measured67.1%Only GPT-5.2 measured
MedQA (Vals AI)Not measured94.1%Only GPT-5.2 measured
MedCode (Vals AI)Not measured49.8%Only GPT-5.2 measured
MMLU Pro (Artificial Analysis run)84.9%87.4%GPT-5.2
SimpleQA VerifiedNot measured37.1%Only GPT-5.2 measured
LMArena Text: science14301425R1 0528
LMArena Text: medicine and healthcare14451425R1 0528
LMArena Text: legal and government14191427GPT-5.2
LMArena Text: business and finance14021414GPT-5.2
Neither model has a result on the other 3 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index16
Web design benchmarks for R1 0528 and GPT-5.2
BenchmarkR1 0528GPT-5.2Leader
Design Arena: websites11491196GPT-5.2
Design Arena: UI components11021190GPT-5.2
Design Arena: game dev10981196GPT-5.2
Design Arena: data viz11821202GPT-5.2
Design Arena: 3D11111079R1 0528
Design Arena: all code categories11391171GPT-5.2
Design Arena: SVG10381137GPT-5.2
Design Arena: ASCII artNot measured1153Only GPT-5.2 measured
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index811
General benchmarks for R1 0528 and GPT-5.2
BenchmarkR1 0528GPT-5.2Leader
SAGE (Vals AI)Not measured49.3%Only GPT-5.2 measured
MedScribe (Vals AI)Not measured84.4%Only GPT-5.2 measured
IFBench39.6%75.4%GPT-5.2
Long context reasoning (AA-LCR)55.7%82.7%GPT-5.2
DTBenchNot measured90.9%Only GPT-5.2 measured
LMCANot measured43.9%Only GPT-5.2 measured
Fiction.LiveBench75.0%Not measuredOnly R1 0528 measured
Creative writing (Lech Mazur)81.9%Not measuredOnly R1 0528 measured
CL-benchNot measured18.2%Only GPT-5.2 measured
Furniture assemblyNot measured38.3%Only GPT-5.2 measured
LMArena Text14281416R1 0528
LMArena Text: instruction following13821409GPT-5.2
LMArena Text: creative writing14051376R1 0528
LMArena VisionNot measured1246Only GPT-5.2 measured
LMArena Text: English14371422R1 0528
LMArena Text: non-English14121405R1 0528
LMArena Text: Chinese14421447GPT-5.2
LMArena Text: German14041444GPT-5.2
LMArena Text: Japanese13911420GPT-5.2
LMArena Text: Korean13601374GPT-5.2
LMArena Text: French14171454GPT-5.2
LMArena Text: Spanish14111413GPT-5.2
LMArena Text: Russian14231415R1 0528
LMArena Text: Polish14281426R1 0528
LMArena Text: multi-turn14051419GPT-5.2
LMArena Text: longer queries13911413GPT-5.2
LMArena Text: writing and language14021392R1 0528
LMArena Text: entertainment and media14001378R1 0528
LMArena Vision: EnglishNot measured1243Only GPT-5.2 measured
LMArena Vision: ChineseNot measured1305Only GPT-5.2 measured
LMArena Vision: reading textNot measured1257Only GPT-5.2 measured
LMArena Vision: humorNot measured1227Only GPT-5.2 measured
LMArena Vision: captioningNot measured1267Only GPT-5.2 measured
LMArena Vision: recognizing thingsNot measured1232Only GPT-5.2 measured
LMArena Vision: creative writingNot measured1225Only GPT-5.2 measured
LMArena DocumentNot measured1405Only GPT-5.2 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 R1 0528 and GPT-5.2
BenchmarkR1 0528GPT-5.2Leader
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

R1 0528$0.0016Fits in one request
GPT-5.2$0.0088Fits in one request

R1 0528 costs less

Repository review

50K input + 3K output tokens

R1 0528$0.031Fits in one request
GPT-5.2$0.130Fits in one request

R1 0528 costs less

Agent loop

200K input + 10K output tokens

R1 0528$0.121Exceeds the context window
GPT-5.2$0.490Fits in one request

R1 0528 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 R1 0528 and GPT-5.2
SpecificationDeepSeek R1 0528OpenAI GPT-5.2
MakerDeepSeekOpenAI
WeightsOpen weightsNo public weights listed
ReleasedMay 28, 2025December 10, 2025
Context window164K400K
Input $/Mtok$0.50$1.75
Output $/Mtok$2.15$14.00
Coding index71.458.9
Independent measurements behind it12
Cost per solved taskNot computable$1.91
SourceOpenRouter listingOpenRouter listing

Questions

Is R1 0528 better than GPT-5.2 for coding?

Pick R1 0528 for coding: it scores 71.4 on the Vibe Coding Index against 58.9 for GPT-5.2, though on 1 benchmark against 2. Pick GPT-5.2 when the context window is larger, 400K against 164K tokens, or when what matters is LiveCodeBench (Artificial Analysis run) or Terminal-Bench Hard (Artificial Analysis run), where it leads.

Which is cheaper, R1 0528 or GPT-5.2?

R1 0528 costs $0.50 per million input tokens and $2.15 per million output tokens; GPT-5.2 costs $1.75 and $14.00. Cost per solved task is known for GPT-5.2 only.

Which has the bigger context window, R1 0528 or GPT-5.2?

GPT-5.2, with 400K tokens against 164K.

How many benchmarks is the R1 0528 vs GPT-5.2 comparison based on?

55 benchmarks have a result for both models as of October 21, 2026. R1 0528's coding index rests on 1 independent benchmark and GPT-5.2's on 2. Where only one model has a result, the page says so instead of naming a winner.

Compare next