Skip to main content

Model comparison

GPT-5.2-Codex vs GPT-5.4

Data verified

Head-to-head evidence from 15 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

59.1/100
Margin
15.1pts
winning →
OpenAI
74.24/100
0 category wins0 category wins

Public leaderboard positions: GPT-5.2-Codex #58 (Supported); GPT-5.4 #8 (Supported). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. GPT-5.2-Codex and GPT-5.4 share 15 comparable benchmark results. 0 of 8 categories are comparable. 0 results are unique to GPT-5.2-Codex; 37 to GPT-5.4.

Updated July 20, 2026
Shared results
15
GPT-5.2-Codex only
0
GPT-5.4 only
37
Comparable categories
0 / 8

Benchmark data for GPT-5.2-Codex and GPT-5.4 is coming soon on BenchLM.

Confidence note. This is a partial-evidence comparison with 15 shared benchmark results across 6 evidence categories; 0 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

BenchLM has partial data for these models, but not enough overlapping benchmark coverage to produce a fair score-level comparison yet.

GPT-5.4 is priced at $2.50 input / $15.00 output per 1M tokens, versus $1.75 input / $14.00 output per 1M tokens for GPT-5.2-Codex. GPT-5.4 has the larger context window at 1.05M, compared with 400K for GPT-5.2-Codex.

Category breakdown

Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.

Category scores and score margins for GPT-5.2-Codex and GPT-5.4
CategoryGPT-5.2-CodexΔGPT-5.4
AgenticGPT-5.2-CodexNot measuredMarginNo overlapGPT-5.477.2
CodingGPT-5.2-CodexNot measuredMarginNo overlapGPT-5.457.7
KnowledgeGPT-5.2-CodexNot measuredMarginNo overlapGPT-5.457.6
MathGPT-5.2-CodexNot measuredMarginNo overlapGPT-5.442.5
MultimodalGPT-5.2-CodexNot measuredMarginNo overlapGPT-5.473.2

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricGPT-5.2-CodexGPT-5.4Comparison
Input / output priceUSD per 1M tokensGPT-5.2-Codex$1.75 input / $14 outputGPT-5.4$2.5 input / $15 outputGPT-5.2-Codex has the lower combined listed price.
Generation speedtokens per secondGPT-5.2-Codex123 tok/sGPT-5.474 tok/sGPT-5.2-Codex has the higher measured throughput.
First-answer latencyseconds to first tokenGPT-5.2-Codex87.34 sGPT-5.4151.79 sGPT-5.2-Codex reaches the first token sooner.
Context windowmaximum listed tokensGPT-5.2-Codex400KGPT-5.41.05MGPT-5.4 lists the larger context window.

Benchmark Deep Dive

Agentic
BenchmarkGPT-5.2-CodexGPT-5.4Result
τ²-bench resultsSource 92.1%87.1%GPT-5.2-Codex leads
Gert LabsSource 51.79%64.89%GPT-5.4 leads
JobBenchSource 26.0%38.9%GPT-5.4 leads
Terminal-Bench 2.0Source 75.1%Not comparable
CyberGymSource 79.0%Not comparable
BrowseCompSource 82.7%Not comparable
OSWorld-VerifiedSource 75%Not comparable
MCP AtlasSource 70.6%Not comparable
ToolathlonSource 54.6%Not comparable
Claw-EvalSource 60.3%Not comparable
DeepSearchQASource 73.6%Not comparable
AA Agentic IndexSource 41.1%Not comparable
APEX-Agents-AASource 33.3%Not comparable
GDPval-AASource 44.7%Not comparable
GDPval-AASource 1395Not comparable
ResearchClawBenchSource 15.3%Not comparable
ExploitGymSource 6.0%Not comparable
Coding
BenchmarkGPT-5.2-CodexGPT-5.4Result
Vibe Code BenchSource 37.91%67.42%GPT-5.4 leads
AA-SciCodeSource 54.6%56.6%GPT-5.4 leads
LiveCodeBench ProSource 87.5%Not comparable
SWE-bench ProSource 57.7%Not comparable
React Native EvalsSource 85.3%Not comparable
AA Coding IndexSource 71.0%Not comparable
Reasoning
BenchmarkGPT-5.2-CodexGPT-5.4Result
AA-LCRSource 75.7%74.0%GPT-5.2-Codex leads
CritPtSource 8.7%23.4%GPT-5.4 leads
Knowledge
BenchmarkGPT-5.2-CodexGPT-5.4Result
Artificial Analysis Intelligence IndexSource 40.1%51.4%GPT-5.4 leads
AA-GPQA DiamondSource 89.9%92.0%GPT-5.4 leads
AA-HLESource 33.5%41.6%GPT-5.4 leads
AA-Omniscience IndexSource -2.5%5.7%GPT-5.4 leads
AA-Omniscience AccuracySource 40.7%50.0%GPT-5.4 leads
AA-Omniscience Hallucination RateSource 72.8%88.6%GPT-5.2-Codex leads
GPQASource 92.8%Not comparable
HLESource 52.1%Not comparable
HLE w/o toolsSource 39.8%Not comparable
GPQA-DSource 92.8%Not comparable
HealthBench HardSource 40.1%Not comparable
MedXpertQA (Text)Source 59.6%Not comparable
HealthBench ProfessionalSource 48.1%Not comparable
Math
BenchmarkGPT-5.2-CodexGPT-5.4Result
FrontierMath v2 (Tiers 1-3)Source 47.600%Not comparable
FrontierMath v2 (Tier 4)Source 27.100%Not comparable
Multimodal
BenchmarkGPT-5.2-CodexGPT-5.4Result
AA-MMMU-ProSource 76.3%78.4%GPT-5.4 leads
MMMU-ProSource 81.2%Not comparable
OfficeQA ProSource 53.2%Not comparable
MMMU-Pro w/ PythonSource 82.1%Not comparable
CharXivSource 82.8%Not comparable
ERQASource 65.4%Not comparable
SimpleVQASource 61.1%Not comparable
ScreenSpot ProSource 85.4%Not comparable
ZeroBenchSource 41.0%Not comparable
MedXpertQA (MM)Source 77.1%Not comparable
Design Arena WebsiteSource 1252Not comparable
Inst. Following
BenchmarkGPT-5.2-CodexGPT-5.4Result
AA-IFBenchSource 77.6%73.9%GPT-5.2-Codex leads
Frequently Asked Questions (3)

Can I compare GPT-5.2-Codex and GPT-5.4 on BenchLM yet?

Not fully yet. BenchLM is tracking both models, but the sourced benchmark breakdown for this comparison is still coming soon.

Why does this comparison show “coming soon”?

BenchLM only shows category winners and benchmark-level calls when we have sourced results that can be compared fairly. For these models, the public benchmark coverage is not complete enough yet.

What data is available for GPT-5.2-Codex and GPT-5.4 today?

GPT-5.2-Codex: $1.75 input / $14.00 output per 1M tokens GPT-5.4: $2.50 input / $15.00 output per 1M tokens Both model pages still include creator, context window, reasoning mode, and other metadata while benchmark coverage fills in.

Related Comparisons

Last updated: July 20, 2026

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.