V8.5 LLM Guide

v8 5 llm quick start guide v8 5 is a hybrid model with built in reasoning capabilities reasoning can take longer to process and may subtly change behavior the default is none reasoning effort slider (speed vs depth) location click the llm version icon in the upper right corner of the homepage the reasoning slider controls how much “thinking time” the ai gets before responding speedy (no reasoning) the ai answers immediately with no extra thinking fastest, but not as detailed moderate gives the ai a bit more time to think more accurate and consistent without slowing things down too much slow the ai can think longer and consider more details before responding better for complex tasks very slow maximum reasoning time best for deep, technical, or nuanced questions, but responses take longer in simple terms, the slower the setting, the more time the ai gets to think, and the more thorough the answer will be baseline setup (when transitioning to v8 5) set dynamism = 0 95 we strongly recommend keeping close to 0 95, going up and down slightly as needed for variation response directive (rd) (optional — keep minimal, put key traits to reinforce or see below for common directives) example message (em) optional you can also put a few simple rules here (formatting & style) example ##response rules #takes a proactive role in plot progression #response structure actions always in asterisks only "speech with quotation marks " pro tips (same core behavior as v8) fewer rules = smoother, more natural output over constraining causes hesitation or repetition too many requirements = “freeze” behaviors the model repeats because it can’t satisfy all your rules at once too verbose or over the limit of desired length set a much lower limit than what you actually need a smaller target provides enough buffer to stay within your preferred range if you want approximately 900 characters, request under 500 use \[impt …] for the rules that matter most this raises priority and reduces drift example rd (optional) \[impt medium response length ] em \[impt use third‑person; actions in asterisks , speech in "quotes" ] give each tweak a few turns to settle before adjusting again note strict concision rules in rd can reduce creativity; em/backstory often works better for pacing