NOUMENOpen app
NOUMEN

OpenAI · Frontier

GPT 5.2

Added Dec 11, 2025

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Use in the app

GPT-5.2 is an OpenAI GPT-5-series model with strong reasoning and long-context performance.

  • Reasoning
  • Tools
  • Vision
  • Structured output
  • PDF input
Input
Text, Images
Output
Text
Reasoning effort
none · low · medium · high · xhigh
Discovery tier
Frontier

Pricing

USD per 1M tokens
Input
$1.75
Output
$14.00
Cache read
$0.18

Reference prices as published by the model provider. API example

Specifications

Context window
400K
Max output
128K
Average output (7 days)
172 tokens
Longer than 17% of 465 models

Side by side

GPT 5.2 compared

Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

GPT 5.2OpenAIThis modelGemma 4 26B A4BGoogleRemove
ProviderOpenAIGoogle
Discovery tierFrontierOpen weights
AddedDec 11, 2025Apr 2, 2026
Limits
Context window400KBest262.14K
Max output128K131.07KBest
Average output172 tokens131 tokens
Pricing, USD per 1M tokens
Input$1.75$0.09Best
Output$14.00$0.30Best
Cache read$0.18$0.05Best
Performance
Throughput—85.5 tok/s
Time to first token—2.3 s
Request success (24h)100.0%100.0%
Benchmark indexes
Intelligence Index17.0Best16.7
Coding Index—39.3
Math Index51.0—
Capabilities
ReasoningYesYes
ToolsYesYes
VisionYesYes
Structured outputYesYes
PDF inputYesNo
Video inputNoYes
Reasoning effortnone, low, medium, high, xhigh—

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

  • GPT 5.2
  • Gemma 4 26B A4B

Intelligence Index

  • GPT 5.217.0
  • Gemma 4 26B A4B16.7

GPT 5.2: Better than 45% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • GPT 5.2: —
    • Gemma 4 26B A4B: 39.3

  • Math Index

    Composite of the math evaluations

    • GPT 5.2: 51.0
    • Gemma 4 26B A4B: —

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • GPT 5.2: —
    • Gemma 4 26B A4B: 1.9%

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • GPT 5.2: —
    • Gemma 4 26B A4B: 0.2%

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • GPT 5.2: —
    • Gemma 4 26B A4B: 315 Elo

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • GPT 5.2: —
    • Gemma 4 26B A4B: 565 Elo

Document reasoning

  • AA-LCRv1.1

    Reasoning across long documents

    • GPT 5.2: 45.7%
    • Gemma 4 26B A4B: 65.7%Top

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • GPT 5.2: 8.0%
    • Gemma 4 26B A4B: 19.3%Top

  • GPQA Diamond

    Graduate-level science questions

    • GPT 5.2: 71.2%
    • Gemma 4 26B A4B: 79.2%Top

  • MMLU-Pro

    Multi-subject knowledge and reasoning

    • GPT 5.2: 81.4%
    • Gemma 4 26B A4B: —

  • IFBench

    Following precise output instructions

    • GPT 5.2: 47.4%
    • Gemma 4 26B A4B: 72.4%Top

Coding

  • SciCode

    Scientific computing in Python

    • GPT 5.2: —
    • Gemma 4 26B A4B: 40.0%

  • LiveCodeBench

    Recent competitive programming problems

    • GPT 5.2: 66.9%
    • Gemma 4 26B A4B: —

Math

  • AIME 2025

    Competition mathematics, 2025 problems

    • GPT 5.2: 51.0%
    • Gemma 4 26B A4B: —

Legacy evaluations

  • Terminal-Bench Hard

    Earlier terminal task suite

    • GPT 5.2: 31.8%Top
    • Gemma 4 26B A4B: 13.6%

  • τ²-Bench Telecom

    Tool use in a customer-support simulation

    • GPT 5.2: 46.5%Top
    • Gemma 4 26B A4B: 43.6%

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • GPT 5.2: 45.7%
    • Gemma 4 26B A4B: 65.7%Top

Scores from Artificial Analysis, updated Oct 26, 2026. Artificial Analysis

Live routing data

Performance and reliability

Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.

Prompt cache hits
48.7%
About $0.98 per 1M input tokens with caching
Request success (24h)
100.0%
Reliable

Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.

Hourly request success · last 72 hoursLowest hour 100.0%
  • Healthy (98% or more)
  • Degraded (90–98%)
  • Down (under 90%)
  • No requests

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart
curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-5.2",
    "reasoning_effort": "medium",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'
import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="openai/gpt-5.2",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

OpenAI

Other OpenAI models, starting with the same family.