GPT-5.6 将 Sol、Terra 与 Luna 整合为一个三阶模型家族,专为前沿智能、高效的日常工作,以及跨技术和专业领域的高并发快速执行而打造。

Model Type:
Pricing: Input 280 credits / 1M tokens (≈ $1.40), Cached Input 28 credits / 1M tokens (≈ $0.14), Cache Writes 350 credits / 1M tokens (≈ $1.75), Output 1680 credits / 1M tokens (≈ $8.40). ~72% cheaper than official pricing. High-tier top-ups (+10% bonus) bring effective pricing down to ~90% of the above.

Automatically append previous messages to maintain multi-turn context. May increase token usage.

Function calling
Web Access
Reasoning effort
Stream Response

How can I help you today?

README

Complete guide to using gpt-5-6-sol

借助 Kie.ai 上的 GPT-5.6 API 扩展极具野心的 AI 产品

Original image
GPT-5.6 Sol API:面向前沿智能
GPT-5.6 Terra API:提供平衡的日常性能
GPT-5.6 Luna API:专为快速且经济的大规模扩展而生
GPT-5.6 API 通过 Max 与 Ultra 提供高效推理能力
ChatGPT-5.6 API:全面提升端到端知识工作与交付物质量
GPT-5.6 Sol API 助力网络安全、科学与 AI 研究
OpenAI API 依托 GPT-5.6 能力扩展安全与防护机制
对比维度GPT-5.6 SolGPT-5.6 TerraGPT-5.6 Luna
Official PositioningFlagship frontier modelBalanced lower-cost model for everyday workFastest and most affordable model
Capability DirectionMaximum intelligence and efficiencyStrong performance with practical cost efficiencyFast execution for high-volume workloads
Core StrengthsAdvanced coding, knowledge work, computer use, cybersecurity, science, and complex reasoningEveryday coding, analysis, professional work, and balanced production tasksFrequent requests, scalable processing, and cost-sensitive workloads
Agents’ Last Exam52.70%50.40%50.30%
Artificial Analysis Coding Agent Index v1.1GPT-5.6 的发布时间是什么时候?77.474.6
SWE-Bench Pro64.60%63.40%62.70%
Terminal-Bench 2.188.80%87.40%84.70%
BrowseComp90.40%87.50%83.30%
OSWorld 2.062.60%50.20%45.60%
SEC-Bench Pro71.20%57.70%48.90%
GeneBench Pro28.70%23.30%10.80%
MRCR v2, 256K–512K91.50%89.60%41.30%
OpenAI Official Input Price$5 per 1M tokens$2.50 per 1M tokens$1 per 1M tokens
OpenAI Official Output Price$30 per 1M tokens$15 per 1M tokens$6 per 1M tokens
评估维度基准测试GPT-5.6 SolGPT-5.5Claude Mythos 5Claude Fable 5Claude Opus 4.8Gemini 3.1 Pro Preview
Professional WorkAgents’ Last Exam52.70%46.90%—40.50%45.20%32.10%
Professional WorkGDPval-AA v21,747.8 Elo1,493.7 Elo—1,759.6 Elo1,600.1 Elo962.3 Elo
General IntelligenceArtificial Analysis Intelligence Index v4.158.954.8—59.955.746.5
Agentic CodingArtificial Analysis Coding Agent Index v1.1GPT-5.6 的发布时间是什么时候?76.4—77.272.542.7
Software EngineeringSWE-Bench Pro64.60%59.40%80.30%80.00%69.20%54.20%
Long-Horizon EngineeringDeepSWE v1.172.70%67.00%—69.70%59.00%11.80%
Terminal CodingTerminal-Bench 2.188.80%85.60%88.00%83.10%78.90%70.70%
Agentic BrowsingBrowseComp90.40%84.40%88.00%—84.30%85.90%
Scientific ReasoningGPQA Diamond94.60%93.60%94.10%92.60%92.00%94.30%
Advanced MathematicsFrontierMath Tier 1–3 v289.00%85.30%—87.00%80.00%59.60%
Tool UseToolathlon58.00%55.60%61.70%61.70%59.90%48.80%
基于 GPT-5.6 API 的高级编码智能体
基于 OpenAI GPT-5.6 API 的产品设计与前端开发
ChatGPT-5.6 API:全面赋能专业知识工作
OpenAI API:深度助力网络安全、前沿科学与 AI 研究