OpenAI has rolled out improvements to GPT-5.6 Sol, its mid-tier model, adding a reasoning slider that lets users control how deeply the model thinks through problems. The update focuses on more targeted responses across ChatGPT.
The company is simultaneously shifting its free tier strategy. Starting next week, free users will gain unlimited text conversations with GPT-5.6 Luna, OpenAI's smallest model. Luna includes a button for extended reasoning, but performance gaps remain substantial compared to larger variants.
This two-pronged move reflects OpenAI's strategy of tiering access by capability. Sol improvements address paid subscribers who want control over reasoning depth without maxing out inference costs. The reasoning slider lets users balance speed against thoroughness, useful for tasks ranging from quick summarization to complex problem-solving.
Luna's expanded free access represents a shift from prior arrangements. Free users previously faced stricter quotas or model limitations. Unlimited text chats lower the barrier to trying ChatGPT, but the tradeoff is clear: Luna significantly underperforms Sol and larger models on reasoning tasks, math, coding, and nuanced analysis.
OpenAI faces competitive pressure from Claude's free tier, which offers decent capabilities through Anthropic's standard model. By promoting Luna to unlimited free use, OpenAI keeps users in its ecosystem while reserving Sol and larger models for paying customers.
The reasoning slider on both models addresses a real use case. Some prompts benefit from extensive chain-of-thought reasoning. Others waste tokens and latency through unnecessary deliberation. User control here reduces frustration and gives advanced users fine-grained options that previously required switching between model tiers.
Luna's limitations remain the real story. Free users get volume but sacrifice quality. For serious work, Luna likely frustrates more than impresses. OpenAI knows this. The strategy banks on users experiencing Luna's constraints