Introducing GPT-5.6: Meet Sol, Terra, and Luna
Introducing GPT-5.6: Meet Sol, Terra, and Luna
OpenAI has unveiled three new models of GPT-5.6 named after the solar system, each with distinct features and pricing.
OpenAI has announced new GPT-5.6 models named after the solar system:
Sol – the largest and most powerful model. It outperforms Mythos on many benchmarks while being half the price. The authors claim that the model uses three times fewer tokens, resulting in an overall cost savings of 5-6 times.
Terra – an intermediate option. It is already three times cheaper than Mythos and confidently surpasses Opus 4.8.
Luna – the cheapest and fastest model, still showing good results.
Currently, there are not many benchmarks available, so we will wait for further measurements. General access is promised in the coming weeks, with the models currently available only to a limited group of testers.
Why it matters
AnalysisThese new models could significantly reduce the costs associated with AI technology usage, making them more accessible for developers and companies. Specifically, the reduction in token usage may enhance efficiency across various applications.
Discuss in community
Share your questions and insights with developers
Related in Ecosystem
Practical guides, core terminology, and agent skills
Part I. AI Agent Architecture: The Foundations
The eleven foundational building blocks of modern autonomous agents: model taxonomy, the agentic runtime loop, advanced prompting, distillation, skills architecture, memory systems, function calling, and FinOps.
Part II. AI Agent Architecture: From Model to Memory and Tools
A deep engineering dive: tokenization, KV-cache, sampling determinism, the training stack, quantization, context engineering, the safety stack, production RAG, ReAct, and multi-agent swarm systems.
Frontier Models
The most powerful class of artificial intelligence at the forefront of global research (Claude 3.7 Sonnet, OpenAI o3/GPT-4.5, Gemini 2.0 Pro), defining the limits of modern reasoning, autonomy, and coding capabilities.
LLM (Large Language Model)
A fundamental class of neural network architectures based on autoregressive transformers, predicting the probabilistic distribution of subsequent tokens and demonstrating emergent properties of abstract reasoning, code synthesis, and logical inference.