GPT-5
ClosedGeneral LLMReasoningTool useVision
400Kcontext
$1.25 / $10per 1M tokens (in / out)
—parameters
62stories in our feed
Specs
hover a value to see its sourceContext window400K
Max output125K
Parameters—
Knowledge cutoff2024-09-30
LicenceProprietary
Commercial useAPI only
InputText · Image
OutputText
Familygpt
API price
USD per 1M tokensInput$1.25
Output$10
Cached input$0.125
No price change recorded since we started tracking.
2026-11-02
US Census Bureau survey shows AI adoption reaching 40% at large firmsSimon Willison’s Weblog
2026-10-22
2026-10-15
研究:多智能体框架多数跑不赢单智能体加 CoT-SCSalesforce AI Research
2026-10-11
Pangram 漏检 79.8% 的 Muse-Glimmer 改写论文摘要rohanpaul_ai
2026-10-10
2026-10-07
2026-10-02
ASAD adaptive debugging agentsarXiv: DeepSeek
2026-09-28
美国作家协会诉OpenAI案解封文件显示高管明知使用盗版图书训练模型SuperTechFans
2026-09-18
2026-09-10
OpenAI 发布 ChatGPT 5.2,提升专业能力与视觉理解Liputan6 Tekno
AgentAudit Framework for AI AgentsarXiv: OpenAI
2026-09-09
2026-09-08
OpenAI's Astra Solves Math ProblemsJustine Moore
2026-09-07
Substrate-Aware AI Agents StudyarXiv: OpenAI
2026-09-04
Agent技术选型实验:Stripe等工具成首选shao__meng
2026-09-03
UTP-Bench Travel Planning BenchmarkarXiv cs.AI
2026-09-01
ECGQuest Benchmark for Language ModelsarXiv: DeepSeek
MIOH Benchmark for Object HallucinationarXiv cs.AI
Boundary Metadata Collapse in Multi-Agent LLM HandoffsarXiv: DeepSeek
2026-08-28
Mentalization in Humans and Language ModelsarXiv: DeepSeek
2026-08-10
2026-08-07
2026-08-05
2026-08-02
Gary Marcus盘点OpenAI Astra八大误解:数学强不等于通用智能Gary Marcus
2026-07-31
2026-07-26
2026-07-14
Claude Fable 5在生物医学挑战问题上的能力评估arXiv: Anthropic
2026-07-13
地缘政治对齐:大型语言模型中的背书效应arXiv: DeepSeek
2026-07-02
2026-06-29
Nature Medicine论文开源测试套件:GPT-5.5 Pro得分79%@koltregaskes
2026-06-28
2026-06-26
美国政府或审批GPT-5.6使用,开源模型意义凸显berryxia