VerdictPal
Join

Models / OpenAI

GPT-5.2 (medium)

GPT-5.2 (medium) · Released 2025-12-11

AA Intelligence Index 38.0. Listed on Artificial Analysis with API pricing $1.75/$14 per 1M tokens. Benchmark rows sync from the AA snapshot — see the ledger on this page.

Frontier model tracked on Artificial Analysis. Use the benchmark ledger below for sourced scores; we do not infer numbers beyond published rows.

#66 of 168 on AA Index · current snapshot
39 AA Index · 12 benchmark rows · 1 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite63
Context128K
Input / 1M$1.8
Output / 1M$14
Knowledge cutoffopen
Statussolid

Family profile

Best published score in each covered benchmark family.

78/100 avg
Knowledge
86
Reasoning
86
Math
97
Coding
89
Agentic
43
Long context
69

6 tested benchmark families

Benchmark placements

Where GPT-5.2 (medium) places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.

BenchmarkFamilyRankScoreSourceDate
AIMEMath#3/ 3797Artificial Analysis2026-09-01
LiveCodeBenchCoding#4/ 3189Artificial Analysis2026-09-01
MMLU-ProKnowledge#20/ 3986Artificial Analysis2026-09-01
IFBenchReasoning#70/ 12765Artificial Analysis2026-09-01
SciCodeCoding#72/ 18046Artificial Analysis2026-09-01
Artificial Analysis Intelligence IndexReasoning#81/ 18739Artificial Analysis2026-09-01
GPQA DiamondReasoning#82/ 18586Artificial Analysis2026-09-01
Humanity's Last ExamReasoning#96/ 18027Artificial Analysis2026-09-01
AA-LCRLong context#97/ 17969Artificial Analysis2026-09-01
Terminal-BenchAgentic#104/ 18443Artificial Analysis2026-09-01
AA time to first tokenPerformance#132/ 186Artificial Analysis2026-09-01
AA output speedPerformance#133/ 187Artificial Analysis2026-09-01
Family context

The three highest-scoring models with pages in each capability family. Where GPT-5.2 (medium) shows up, it's highlighted.

KnowledgeGPT-5.2 (medium) · 86
  1. 1
    GPT 5.5
    91
  2. 2
    Claude Opus 4.8
    91
  3. 3
    Claude Sonnet 4.6
    91
ReasoningGPT-5.2 (medium) · 86
  1. 1
    Claude Mythos Preview
    95
  2. 2
    Gemini 3.7 Flash
    95
  3. 3
    Grok 4.6
    95
MathGPT-5.2 (medium) · 97
  1. 1
    OpenAI o3
    99
  2. 2
    GPT-5 (medium)
    99
  3. 3
    GPT-5 (high)
    99
CodingGPT-5.2 (medium) · 89
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticGPT-5.2 (medium) · 43
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextGPT-5.2 (medium) · 69
  1. 1
    Muse Spark 1.2
    83
  2. 2
    Kimi K3
    83
  3. 3
    Gemini 3.5 Flash
    81
Newest receipts

Benchmark rows added to the public ledger for GPT-5.2 (medium) in the last 120 days. Older rows live in the full table below.

12 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context69%Artificial Analysis5
AA output speedPerformance0 t/sArtificial Analysis5
AA time to first tokenPerformance0.00sArtificial Analysis5
AIMEMath96.67%Artificial Analysis5
Artificial Analysis Intelligence IndexReasoning38.9%Artificial Analysis5
GPQA DiamondReasoning86.4%Artificial Analysis5
Humanity's Last ExamReasoning26.7%Artificial Analysis5
IFBenchReasoning65.24%Artificial Analysis5
LiveCodeBenchCoding89.4%Artificial Analysis5
MMLU-ProKnowledge85.9%Artificial Analysis5
SciCodeCoding46.2%Artificial Analysis5
Terminal-BenchAgentic43.18%Artificial Analysis5
Family coverage

A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.

Knowledge86
Reasoning86
Math97
Coding89
Agentic43
Long context69
Price vs. performance
0255075100$0.10$0.30$1$3$10$30input USD / 1M tokens · log scale →← intelligence

Hover any dot for name, score, and input price. Keyboard: tab through the top twelve, or use the ranking below.

AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotMetaMiniMaxxAIZhipuXiaomiOtherMeituanCohereNVIDIA

Full benchmark ledger

Every catalog benchmark for GPT-5.2 (medium). Scores link to the original source; gaps mean no public row exists yet.

12 sourced rows

12 of 37 catalog benchmarks have a sourced row for GPT-5.2 (medium).

Knowledge

Reasoning

Math

Coding

Agentic

Long context

Performance

Sources