OpenAI unveils GPT-5.6 Work model: New AI tiers Sol, Terra & Luna

Quick summary – click for full details

Concise summary of key highlights

In one line

OpenAI launches GPT-5.6-powered ChatGPT Work model with advanced AI capabilities and competitive benchmarks.

Key points

• New GPT Work series

OpenAI introduces the GPT Work series, powered by the latest GPT-5.6 model, offering advanced AI capabilities across multiple domains.

• Three performance tiers

The GPT-5.6 model is available in three tiers—Sol, Terra, and Luna—each offering varying levels of performance and cost efficiency.

• Superior benchmark performance

GPT-5.6 Sol outperforms Anthropic’s Fable 5 series in benchmark tests, including ALE and Artificial Analysis Coding Agent Index.

• Unified desktop app

OpenAI launches a new ChatGPT desktop app integrating Chat, Work, and Codex for diverse user needs.

• Global rollout phase

The GPT-5.6 model is gradually rolling out globally, with full availability expected by the weekend.

Key statistics

53.6

ALE benchmark score for GPT-5.6 Sol

13.1 points

ALE benchmark lead over Fable 5

80.0

Artificial Analysis Coding Agent Index score for GPT-5.6 Sol

2.8 points

Coding benchmark lead over Fable 5

Processed with AI. Reviewed by DH Digital Team.

Published 10 July 2026, 08:03 IST