OpenAI Officially Launches Latest Generation Large Language Model Series GPT-5.6

OpenAI officially released the latest GPT-5.6 series models on Thursday, injecting another heavyweight product into the increasingly crowded AI market.

GPT-5.6 is divided into three versions: Sol (main version), Terra (mid-tier), and Luna (economy version). These three models expand user scenarios across multiple fields, with the company promising strong capabilities in enterprise office work, code development, and even scientific research.

CEO Sam Altman stated that the new series models have significantly improved efficiency and cost compared to the previous generation. He recently revealed in an interview with CNBC that Sol improved token efficiency by 54% in AI programming tasks.

Notably, OpenAI refers to GPT-5.6 as the "strongest cybersecurity model to date," claiming it "achieves frontier-level performance with significantly fewer tokens."

The model's capabilities in the cybersecurity field are closely watched. Previously, the Trump administration attempted to limit the launch of the model due to concerns about potential misuse. GPT-5.6 supports various defensive security activities, including threat modeling, code review and vulnerability patching, and blue team exercises (i.e., simulating attacks on its own systems to identify potential vulnerabilities in advance).

Meanwhile, OpenAI also released a new tool called ChatGPT Work, designed specifically for enterprise teams, supporting desktop, web, and mobile operation, assisting with daily tasks such as drafting documents, creating spreadsheets, and presentations.

The launch of the GPT-5.6 series follows similar moves by competitors SpaceXAI and Meta this week.

However, GPT-5.6 and its marketing strategy seem to target OpenAI's main competitor, Anthropic. Anthropic has established a friendly "low-key challenger" image in the AI race by consistently focusing on enterprise clients and has won the support of more and more users.

Facing this competitive landscape, OpenAI cited the well-known industry evaluation metric "Artificial Analysis Programming Agent Index," claiming its latest series models comprehensively surpass Anthropic's models across all metrics.

OpenAI positions Sol as the "strongest programming model to date" and directly compares it with Anthropic's recently launched Fable. According to the Programming Agent Index, OpenAI states "Sol sets a new industry benchmark with a score of 80, 2.8 points higher than Fable 5, while using less than half the output tokens, taking less than half the time, and costing approximately one-third less."

OpenAI added: "This advantage is reflected across the entire series: Terra scores slightly higher than Fable 5, while Luna surpasses Opus 4.8."

Currently, GPT-5.6 is available on ChatGPT, Codex, and the OpenAI API. Billed per million tokens, the specific prices are as follows: Sol at $5 input/$30 output, Terra at $2.50 input/$15 output, and Luna at $1 input/$6 output.

Q&A

Q1: What versions does GPT-5.6 have, and who are they suitable for?

A: GPT-5.6 is divided into three versions: Sol is the main version with the strongest performance, suitable for enterprise users with high requirements for programming and cybersecurity; Terra is the mid-tier, balancing performance and cost; Luna is the economy version, suitable for users with limited budgets. All three models are available on ChatGPT, Codex, and the OpenAI API, billed per million tokens.

Q2: What specific capabilities does GPT-5.6 have in cybersecurity?

A: OpenAI positions GPT-5.6 as the "strongest cybersecurity model to date." It supports various defensive security activities, including threat modeling, code review and vulnerability patching, and blue team exercises, which simulate hacker attacks on its own systems to help users identify and fix security vulnerabilities before real attacks occur.

Q3: What is ChatGPT Work, and what functions does it have?

A: ChatGPT Work is an enterprise office tool released by OpenAI alongside GPT-5.6, designed specifically for enterprise teams, supporting desktop, web, and mobile use. It can assist users with daily tasks, such as drafting documents, creating spreadsheets, and presentations, positioning itself as an intelligent office assistant for enterprise teams.

This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.