Perplexity Agent
OpenAI's April API update ships GPT-5.1 as the new default reasoning engine, with Codex variants and GPT-5.4 mini and nano tightening the grip on agentic
Model details
OpenAI framed GPT-5.1 as an upgrade to the GPT-5 series that pairs a warmer, more conversational default mode (Instant) with a more persistent Thinking variant for complex reasoning, explicitly noting improvements in both raw intelligence and communication style. The Instant branch is positioned as the everyday workhorse, designed to feel playful yet clear and useful, while Thinking is aimed at multi-step problems where extended analysis pays off. This bifurcated design reflects a deliberate split between response speed and reasoning depth rather than asking users to pick a single model for every task.
In practice, developers can tune how hard the model thinks using a `reasoning_effort` parameter with five levels (none through xhigh), which changes both the latency profile and how persistent the model is on harder problems. The release also introduced seven personality presets that shape tone without affecting underlying capability, and reported meaningful speedups on simple requests compared with the previous generation. Independent third-party studies have already begun evaluating GPT-5.1 against predecessors like GPT-4o on specialized exams such as Persian-language rheumatology board questions, and third-party leaderboards have since recorded other models surpassing GPT-5.1-High in specific math benchmarks, signaling an active competitive landscape rather than a static state-of-the-art position.
Perplexity Agent
OpenAI's April API update ships GPT-5.1 as the new default reasoning engine, with Codex variants and GPT-5.4 mini and nano tightening the grip on agentic
Perplexity Agent
Editor’s note (April 15, 2026): We updated this post to include information about the deprecation of GPT 5.1. We have deprecated the following models across all GitHub Copilot experiences (including Copilot Chat, inline edits, ask and agent modes, and code completions) on April 1, 2026. | Model | Deprecation date | Sug
Perplexity Agent
Large language models are increasingly integrated into medical education, yet their performance in non-English clinical examinations, particularly Persian, remains limited. This study evaluated how GPT-4o and GPT-5.1 perform on Iranian Rheumatology Board examination questions. A total of 204 multiple-choice items were
Perplexity Agent
Baidu's ERNIE-5.0-0110 ranks #8 globally on LMArena, becoming the only Chinese model in the top 10 while outperforming GPT-5.1-High.
Perplexity Agent
Baidu's ERNIE-5.0-0110 ranks #8 globally on LMArena, becoming the only Chinese model in the top 10 while outperforming GPT-5.1-High.
Perplexity Agent
GLM-4.7 is an open-weights model with benchmark parity with GPT-5.1 and Claude Sonnet 4.5.