Tags: token efficiency

OpenAI has released GPT-5.4, positioning it as its most capable model for professional work. The update focuses on productivity tasks, delivering a notable jump in spreadsheet accuracy and speed, an Excel sidebar add‑on, and stronger presentation output. Human reviewers reported fewer errors and false claims, while token efficiency improvements enable longer, more complex workflows. The model is available to Pro and Enterprise users through ChatGPT and the API, with distinct pricing tiers for standard and Pro versions. Ler mais

OpenAI announced the release of GPT-5.4, its newest foundation model designed for professional workloads. The model is offered in three versions—a standard release, a high‑performance Pro edition, and a reasoning‑focused Thinking edition. GPT-5.4 features a context window of up to one million tokens and delivers significant token‑efficiency gains, allowing it to solve tasks with fewer tokens than prior models. Benchmark scores show record performance across computer‑use and knowledge‑work tests, while safety updates cut hallucinations by roughly one‑third. A new tool‑calling architecture called Tool Search reduces token overhead when accessing many tools, and a safety evaluation demonstrates lower risk of deceptive chain‑of‑thought behavior in the Thinking version. Ler mais

AI coding agents face limits on the amount of code they can process at once, which can quickly consume token or usage limits when large files are fed directly into a language model. To work around these constraints, developers fine‑tune models to generate auxiliary scripts that extract needed data, allowing the agents to operate on smaller, targeted inputs. Techniques such as dynamic context management and context compression let agents summarize past interactions, preserving essential details while discarding redundant information. These approaches enable semi‑autonomous tools like Claude Code and OpenAI Codex to handle complex codebases more efficiently without overwhelming the underlying model. Ler mais