LogoLMSpeed
Back to models

Model compare

GPT-5.2 vs GPT-5.3 Codex

The readout for GPT-5.2 and GPT-5.3 Codex, before the detailed comparison sheet.

Model A

GPT-5.2

OpenAI

Contender
vs

Model B

GPT-5.3 Codex

OpenAI

Leading

Key Takeaways

Weighted outcome: GPT-5.3 Codex. Benchmark capability categories carry 80%, while price, API performance, and availability carry 20%.

Decision read

GPT-5.3 Codex

GPT-5.3 Codex has the higher weighted result; Model A / B score 18.7 to 81.3.

Evidence depth

122 data points

Includes 29 benchmark rows, 0 audit samples, and 10 provider examples.

Selection signal

Start with GPT-5.3 Codex

The charts below split 39 high-signal samples across speed, scores, and audit health.

Change comparison

Switch either side of this report to compare another model with the same LMSpeed data pipeline.

Select a different model to open a new comparison URL.

Comparison sheet

This report only uses LMSpeed data for GPT-5.2 and GPT-5.3 Codex: pricing, speed aggregates, third-party benchmark scores, and shared provider samples.

Model compare
GPT-5.2
GPT-5.3 Codex
Overall leaderContenderLeading
Weighted overall score18.7 pts81.3 pts
Benchmark category leads0 categories5 categories
Operational advantagesCheapest input price, First-token latency, Free providersAverage speed, Provider coverage
Context window400K tokens400K tokens
Max output128K tokens128K tokens
Modalities

Input

FileImageText

Output

Text

Input

TextImageFile

Output

Text
Features
Text inputImage inputFile inputText outputTool callingStructured outputsJSON modeReasoning
Text inputImage inputFile inputText outputTool callingStructured outputsJSON modeReasoning

The overall result weights benchmark capability categories at 80% and price, API speed/latency, and availability at 20%. Recent test volume does not affect the winner, and missing benchmark categories are excluded.

Model metadata

Model compare
GPT-5.2
GPT-5.3 Codex
DeveloperOpenAIOpenAI
ReleasedDec 2025Feb 2026
ParametersNo dataNo data
TokenizerGPTGPT
Knowledge cutoffNo dataNo data
OpenRouter IDopenai/gpt-5.2openai/gpt-5.3-codex
ReferencesNo dataNo data

When to choose each model

This report only uses LMSpeed data for GPT-5.2 and GPT-5.3 Codex: pricing, speed aggregates, third-party benchmark scores, and shared provider samples.

GPT-5.2

GPT-5.2 has these operational advantages: Cheapest input price, First-token latency, Free providers.

GPT-5.3 Codex

GPT-5.3 Codex is stronger in benchmark categories (Agents, Coding, Reasoning, Knowledge, Multimodal) and operational dimensions (Average speed, Provider coverage).

Benchmark score comparison

Third-party benchmark profile synced into LMSpeed; only metrics available for both models are shown.

Category performance

Compare benchmark category scores on a 0-100 scale. Select a category to inspect the gap.

Model A coverage
7 / 8
Model B coverage
6 / 8
Shared
6 shared categories

Avg. score

GPT-5.2

51.8

Avg. score

GPT-5.3 Codex

55.3

Agents

GPT-5.3 Codex leads by 5.1

GPT-5.247
GPT-5.3 Codex52.1

Coding

GPT-5.3 Codex leads by 0.8

GPT-5.256.3
GPT-5.3 Codex57.1

Reasoning

GPT-5.3 Codex leads by 6.7

GPT-5.253.1
GPT-5.3 Codex59.8

Knowledge

GPT-5.3 Codex leads by 2.7

GPT-5.254.4
GPT-5.3 Codex57.1

Math

GPT-5.2

GPT-5.255.3
GPT-5.3 Codex-

Multilingual

No data

GPT-5.2-
GPT-5.3 Codex-

Multimodal

GPT-5.3 Codex leads by 8.6

GPT-5.242.2
GPT-5.3 Codex50.8

Instruction following

Tied

GPT-5.254.6
GPT-5.3 Codex54.6

Professional benchmark details

Metric-level scores with benchmark source, rank depth, confidence, error, and evaluation date where available.

Aggregatereported

BenchLM overall score

SourceGPT-5.2

GPT-5.2

51.0

Rank #78/111 · confidence 4 · eval date 2025-12-11

+17.0

GPT-5.3 Codex

winner

68.0

Rank #26/111 · confidence 3 · eval date 2026-02-05

Pricingverified

Input price

SourceGPT-5.2 (Non-reasoning)

GPT-5.2

$1.75/M

Rank #140/186 · confidence 4

$0.0000/M

GPT-5.3 Codex

$1.75/M

Rank #140/186 · confidence 4

Pricingverified

Blended price

SourceGPT-5.2 (Non-reasoning)

GPT-5.2

$4.81/M

Rank #155/186 · confidence 4

$0.0000/M

GPT-5.3 Codex

$4.81/M

Rank #155/186 · confidence 4

Pricingverified

Output price

SourceGPT-5.2 (Non-reasoning)

GPT-5.2

$14.00/M

Rank #156/186 · confidence 4

$0.0000/M

GPT-5.3 Codex

$14.00/M

Rank #156/186 · confidence 4

Reasoningreported

BenchLM Reasoning score

SourceGPT-5.2

GPT-5.2

53.7

Rank #55/67 · confidence 4 · eval date 2025-12-11

+24.6

GPT-5.3 Codex

winner

78.3

Rank #5/67 · confidence 3 · eval date 2026-02-05

Reasoningreported

AA-LCR

SourceGPT-5.2

GPT-5.2

82.7

Rank #11/106 · confidence 4 · eval date 2025-12-11

+0.6

GPT-5.3 Codex

winner

83.3

Rank #6/106 · confidence 3 · eval date 2026-02-05

Reasoningverified

GPQA

SourceGPT-5.2 (Non-reasoning)

GPT-5.2

71.2%

Rank #138/218 · confidence 4

+20.3%

GPT-5.3 Codex

winner

91.5%

Rank #21/218 · confidence 4

Reasoningreported

CritPt

SourceGPT-5.2

GPT-5.2

11.6

Rank #31/88 · confidence 4 · eval date 2025-12-11

+5.3

GPT-5.3 Codex

winner

16.9

Rank #22/88 · confidence 3 · eval date 2026-02-05

Reasoningverified

HLE

SourceGPT-5.2 (Non-reasoning)

GPT-5.2

8.0%

Rank #134/216 · confidence 4

+34.5%

GPT-5.3 Codex

winner

42.5%

Rank #22/216 · confidence 4

Codingreported

Vibe Code Bench

SourceGPT-5.2

GPT-5.2

53.5

Rank #6/33 · confidence 4 · eval date 2025-12-11

+8.3

GPT-5.3 Codex

winner

61.8

Rank #4/33 · confidence 3 · eval date 2026-02-05

Codingreported

SWE-bench Verified

SourceGPT-5.2

GPT-5.2

80.0

Rank #14/49 · confidence 4 · eval date 2025-12-11

+5.0

GPT-5.3 Codex

winner

85.0

Rank #6/49 · confidence 3 · eval date 2026-02-05

Codingreported

AA-SciCode

SourceGPT-5.2

GPT-5.2

52.1

Rank #28/113 · confidence 4 · eval date 2025-12-11

+1.1

GPT-5.3 Codex

winner

53.2

Rank #26/113 · confidence 3 · eval date 2026-02-05

Codingreported

SWE-bench Pro

SourceGPT-5.2

GPT-5.2

55.6

Rank #33/50 · confidence 4 · eval date 2025-12-11

+1.2

GPT-5.3 Codex

winner

56.8

Rank #26/50 · confidence 3 · eval date 2026-02-05

Codingreported

BenchLM Coding score

SourceGPT-5.2

GPT-5.2

46.5

Rank #64/87 · confidence 4 · eval date 2025-12-11

+5.5

GPT-5.3 Codex

winner

52.0

Rank #50/87 · confidence 3 · eval date 2026-02-05

Knowledgereported

AA-Omniscience Accuracy

SourceGPT-5.2

GPT-5.2

44.3

Rank #27/106 · confidence 4 · eval date 2025-12-11

+8.6

GPT-5.3 Codex

winner

52.9

Rank #11/106 · confidence 3 · eval date 2026-02-05

Knowledgereported

BenchLM Knowledge score

SourceGPT-5.2

GPT-5.2

77.2

Rank #23/83 · confidence 4 · eval date 2025-12-11

+2.3

GPT-5.3 Codex

winner

79.5

Rank #19/83 · confidence 3 · eval date 2026-02-05

Knowledgereported

AA-HLE

SourceGPT-5.2

GPT-5.2

37.7

Rank #33/114 · confidence 4 · eval date 2025-12-11

+4.8

GPT-5.3 Codex

winner

42.5

Rank #21/114 · confidence 3 · eval date 2026-02-05

Knowledgereported

AA-Omniscience Hallucination Rate

SourceGPT-5.2

GPT-5.2

81.2

Rank #40/106 · confidence 4 · eval date 2025-12-11

+8.0

GPT-5.3 Codex

winner

89.2

Rank #21/106 · confidence 3 · eval date 2026-02-05

Knowledgereported

AA-GPQA Diamond

SourceGPT-5.2

GPT-5.2

90.3

Rank #32/113 · confidence 4 · eval date 2025-12-11

+1.2

GPT-5.3 Codex

winner

91.5

Rank #23/113 · confidence 3 · eval date 2026-02-05

Knowledgereported

Artificial Analysis Intelligence Index

SourceGPT-5.2

GPT-5.2

30.4

Rank #47/117 · confidence 4 · eval date 2025-12-11

+2.1

GPT-5.3 Codex

winner

32.5

Rank #43/117 · confidence 3 · eval date 2026-02-05

Multimodalreported

BenchLM Multimodal Grounded score

SourceGPT-5.2

GPT-5.2

66.3

Rank #31/56 · confidence 4 · eval date 2025-12-11

+9.3

GPT-5.3 Codex

winner

75.6

Rank #20/56 · confidence 3 · eval date 2026-02-05

Multimodalreported

Design Arena Website

SourceGPT-5.2

GPT-5.2

winner

1209.0

Rank #47/78 · confidence 4 · eval date 2025-12-11

+32.0

GPT-5.3 Codex

1177.0

Rank #59/78 · confidence 3 · eval date 2026-02-05

Instruction followingreported

BenchLM Instruction Following score

SourceGPT-5.2

GPT-5.2

92.6

Rank #10/52 · confidence 4 · eval date 2025-12-11

0.0

GPT-5.3 Codex

92.6

Rank #10/52 · confidence 3 · eval date 2026-02-05

Instruction followingreported

AA-IFBench

SourceGPT-5.2

GPT-5.2

75.4

Rank #18/84 · confidence 4 · eval date 2025-12-11

0.0

GPT-5.3 Codex

75.4

Rank #18/84 · confidence 3 · eval date 2026-02-05

Agentsreported

JobBench

SourceGPT-5.2

GPT-5.2

winner

34.3

Rank #10/24 · confidence 4 · eval date 2025-12-11

+0.6

GPT-5.3 Codex

33.7

Rank #11/24 · confidence 3 · eval date 2026-02-05

Agentsreported

Gert Labs

SourceGPT-5.2

GPT-5.2

46.5

Rank #28/49 · confidence 4 · eval date 2025-12-11

+10.9

GPT-5.3 Codex

winner

57.5

Rank #13/49 · confidence 3 · eval date 2026-02-05

Agentsreported

OSWorld-Verified

SourceGPT-5.2

GPT-5.2

47.3

Rank #26/27 · confidence 4 · eval date 2025-12-11

+17.4

GPT-5.3 Codex

winner

64.7

Rank #21/27 · confidence 3 · eval date 2026-02-05

Agentsreported

BenchLM Agentic score

SourceGPT-5.2

GPT-5.2

23.1

Rank #71/76 · confidence 4 · eval date 2025-12-11

+39.7

GPT-5.3 Codex

winner

62.8

Rank #37/76 · confidence 3 · eval date 2026-02-05

Agentsreported

τ²-bench results

SourceGPT-5.2

GPT-5.2

84.8

Rank #46/82 · confidence 4 · eval date 2025-12-11

+1.2

GPT-5.3 Codex

winner

86.0

Rank #42/82 · confidence 3 · eval date 2026-02-05

API audit comparison

Latest completed audits from shared providers, with four safety and integrity score groups plus report links.

Provider
GPT-5.2
GPT-5.3 Codex
No completed audits are available from shared providers yet.

Provider examples

Speed aggregates and input/output pricing share each provider row for real API selection and migration cost checks.

Provider
GPT-5.2
GPT-5.3 Codex
星见雅 API50 tests

GPT-5.2

speed / latency

52 tok/s / 2671ms

input / output

No data

GPT-5.3 Codex

speed / latency

38 tok/s / 4553ms

input / output

No data

N1N15 tests

GPT-5.2

speed / latency

56 tok/s / 1798ms

input / output

No data

GPT-5.3 Codex

speed / latency

N/A / N/A

input / output

No data

小水管 API15 tests

GPT-5.2

speed / latency

71 tok/s / 2893ms

input / output

No data

GPT-5.3 Codex

speed / latency

201 tok/s / 2135ms

input / output

No data

Mars HK14 tests

GPT-5.2

speed / latency

N/A / N/A

input / output

No data

GPT-5.3 Codex

speed / latency

252 tok/s / 2196ms

input / output

No data

A310 tests

GPT-5.2

speed / latency

45 tok/s / 1716ms

input / output

No data

GPT-5.3 Codex

speed / latency

46 tok/s / 2039ms

input / output

No data

180txt API

GPT-5.2

gpt-5.2-chat-latest

speed / latency

No data

input / output

$0/M/$0/M

GPT-5.3 Codex

gpt-5.3-codex-spark

speed / latency

No data

input / output

$0.263/M/$2.10/M

兔子API

GPT-5.2

gpt-5.2-2025-12-11

speed / latency

No data

input / output

$0/request

GPT-5.3 Codex

gpt-5.3-codex-xhigh

speed / latency

No data

input / output

$0/request

Yun API

GPT-5.2

gpt-5.2

speed / latency

No data

input / output

$0.0014/request

GPT-5.3 Codex

gpt-5.3-codex-spark

speed / latency

No data

input / output

$0.060/M/$0.479/M

KFCV50

GPT-5.2

gpt-5.2

speed / latency

No data

input / output

$0.0068/request

GPT-5.3 Codex

gpt-5.3-codex

speed / latency

No data

input / output

$0.0068/request

Newagiai

GPT-5.2

gpt-5.2-all

speed / latency

No data

input / output

$0.0068/request

GPT-5.3 Codex

gpt-5.3-codex-medium

speed / latency

No data

input / output

$0.240/M/$1.92/M

FAQ

Weighted outcome: GPT-5.3 Codex. Benchmark capability categories carry 80%, while price, API performance, and availability carry 20%.

Why is this comparison indexable?
It has 6 verifiable comparison points, and both models have pricing or benchmark data.
Are missing metrics invented?
No. Metrics without LMSpeed data are omitted from this report.

Rankings are based on community-submitted tests and periodic health probes. Advisory only, not official data.