GPT-5.6 Sol Uses Twice the Tokens of GPT-5.5

OpenAI's GPT-5.6 Sol xhigh uses 16.45M tokens per session, 2.25 times the 7.30M tokens per session of GPT-5.5 xhigh, according to a developer's analysis of 1,715 sessions over 14 days. Total token usage rose from 12.17B to 28.22B, a 2.32x increase, effectively halving the value of token-based quotas and draining subscriptions faster.

Codex CLI Has Responsive Tables in the Terminal LLMs use Markdown tables all the time. A few months ago, Codex CLI displayed them as ordinary wrapped text. In a narrow… GPT-5.6 Sol xhigh now uses more than twice as many tokens per session as GPT-5.5 xhigh in my Codex workflow. For Codex users, tokens per session means the total token count divided by the number of sessions. I compared two equal 14-day windows from deduplicated local logs. The sessions were different, but 1,667 sessions before and 1,715 after are similar enough to show why the new model drains my subscriptions. Like-for-like, equal 14-day windows: | Model / effort | Sessions | Tokens | Tokens per session | |---|---|---|---| | GPT-5.5 xhigh | 1,667 | 12.17B | 7.30M | | GPT-5.6 Sol xhigh | 1,715 | 28.22B | 16.45M | The session count rose by only 2.9%. Tokens per session rose by 125%. GPT-5.6 Sol xhigh used 16.45M tokens per session, compared with 7.30M for GPT-5.5 xhigh. That is 2.25x as many tokens for each session. Total usage rose from 12.17B tokens to 28.22B, a 2.32x increase. Roughly the same number of sessions now consumes more than twice the usage allocation. That cuts the effective value of a token-based quota by more than half. My three subscriptions used to survive roughly a week of extreme work. Since GPT-5.6 Sol became the top xhigh model, about one day of medium work can drain them. I also had three to four banked usage resets on each account. I rarely needed them before. Now they are all gone. OpenAI lists the same fresh-input, cached-input, and output token prices https://platform.openai.com/docs/pricing for GPT-5.5 and GPT-5.6 Sol. At the same token price, 2.25x the tokens means roughly 2.25x the cost for a similar token mix. GPT-5.6 Sol also adds a cache-write charge that GPT-5.5 did not have. Give Vroni a GitHub issue, bug report, spec, or rough idea. It reads the repo, plans the change, writes code, runs checks, and works toward a review-ready pull request. Take a look at vroni.com