The AI’s I use like Claude, Deepseek, Grok, and Chat GPT all write in a away that isn’t actual story writing. They usually summarize thing’s, have character’s be omniscient and know past or future events they shouldn’t know, always talk expectantly, give unnecessary sensory and environmental details and it definitely can’t set up a mystery or write a specific genre like comedy. They keep over clarifying as well what something isn’t or is along with even saying what I said not add directly.
I need AI that are at least capable of writing more humanly and following what you tell it. Idk why there isn’t any actual AI built for writing , story writing not essay or school writing. Most can critic something pretty well but can never implement those fixes they suggest. I usually upload my fic through PDF for the AI read . It gets things of what chapter is what at times wrong or completely be unable to remember a chapter ending or the actual one.


You may need to be very specific about how you prompt to get decent results in a particular style. Any wiggle room for ambiguity and the bots’ll drive a bulldozer through what you meant with insane interpretations.
I don’t use cloud based LLMs at all – local only – so I’m not sure what they give you to work with there, but in my own usage, setting the system prompt with clearly defined ROLE, CONSTRAINTS, and other other details has been critical for getting them to do anything useful instead of spitting out generic analysis slop. (Even so, be mindful of context limits.)
LLMs aren’t magic and don’t have infinite memory. In fact, the amount they can hold in their heads effectively is pretty small and you must manage context carefully if you don’t want them to get lost going down irrelevant paths.
What they are very good at is style transfer. If you have text and need it reworded, they can do that trivially.
They are not very good at dealing with large structural reasoning. You really have to break things down and hand-hold them if you want to get them to do stuff like that.
I haven’t done much directly with creative writing in natural language (I’ve mostly done code gen, critical news analysis, image analysis, and a lot of context experimentation) – but the path I’d try out if I wanted to get into that is actually using a lot of writing “self-help” strategies. e.g. Snowflake Method. Lots of outlines. Character profile drafting. Scene and sequel. Apply Vonnegut’s rules for writing. Etc. Then set up multiple LLMs as test readers tuned to different aspects of the writing that you want – like character consistency, diction, appeal to your audience, etc. and have them review it and iterate.
I have everything set up, fanfic is on AO3 and i have chapter 4 out, but the AI’s for the life of me can not listen even when I use specific prompts meant to allow them follow instructions better.
I have a specific type of writing style and formating that really isn’t into using fragmented punchy sentences, over clarification, literary, abstract, poetic and flowery writing, “files away” , “cold tea”, omniscient dialogue or thoughts or narration, summary narration, expectant Dialogue, and etc. If I’m going to edit and write certain things myself, I’d prefer if the AI had a decent base, to work with.
##Writing Guide
TENSE:
PERSPECTIVE & POV:
MECHANICS:
VOICE & STYLE:
SHOW VS TELL:
FORMATTING:
BANNED PHRASES & PATTERNS:
CHARACTERIZATION:
MYSTERY & RESTRAINT:
DIALOGUE and Thoughts:
NARRATIVE LOGIC:
PUNCTUATION & SPEECH PATTERNS:
Ellipses (…): Use for trailing off, pausing, straining, or moments of hesitation where a character struggles to finish a sentence. It should feel natural to the character’s current emotional state. Look up how eclipses are used for other things as well.
Em Dash (—): Use strictly for sharp cutoffs or sudden interruptions. A character can cut themselves off due to various things if another may interrupt ,look at any provided files or example to look at how it’s handled.
Stuttering: Use dashes (B-b-but) only when the character is genuinely overwhelmed, nervous, anxious or physically struggling.
Mouth Sounds: Include “hmm,” “tch,” “huff,” or groans, among other different mouth sounds within dialogue or as brief beats only when they fit the character’s personality and the moment’s intensity or situation. They also should be used appropriately for what each sound is meant to be and be a response to or show of something.
PUNCTUATION & VOLUME PROTOCOL:
Exclamation Marks (!): Use for raised voices or emphasis. One is usually enough for a raised voice.
ALL CAPS: Use strictly for full-volume screaming or yelling.
Volume Variation: Do not use ALL CAPS for low shouts or moments where a voice is raised but the emotion is controlled. Use a single exclamation mark and an action beat instead.
Combined Marks (Interrobangs): Use !? or ?! for shouted questions where the character is loud and confused/panicked.
Double Marks: Use double question marks (??) or exclamation marks (!!) to show extreme confusion or high-intensity emotion, even if the character isn’t shouting.
Letter Repetition: You are permitted to repeat letters to show a character dragging out a word or sound due to physical strain, hesitation, or confusion or other things that may cause it if a situation is that big of those things (e.g., “Ugggh,” “Wait…” “Huuuh??”).
Ellipses (…) vs. Em Dash (—): - Use (…) for trailing off, losing consciousness, or hesitant pauses and other things that fit it being there, like speaking in uncertainty so narration may have this at times.
If you are willing to put all this work into prompting, wouldn’t it be a more gratifying process to just write the thing yourself and save a lot of energy, water , and frustration in the process?
Looking at your writing guide for the LLM, I think my advice about changing workflow was right.
Break the problem down into much smaller steps and then have the LLM do editing passes and self-review from multiple perspectives after getting a crappy rough draft back instead of trying to get it to match your style one-shot.
Using something like this to provide context for the scene when generating a rough draft will likely be helpful:
[SCENE STATE] - POV: [Name] - Location/Time: - Known to POV: - Unresolved/Secrets: - Active Props/Details: - Goal/Conflict:Also, are you dumping in the full PDF of the entire work so far? Try doing a pass to skeletonize it down into an outline – maybe chapter by chapter or even scene by scene – along with getting organized about characters by making profile summaries so that the LLM can understand your story in an already partially digested form instead of trying to do everything all together at once with the huge amount of context reading a full story implies. It’s probably getting lost in the weeds.
I appreciate the advice help. Idk if I can really get them to work that well. Or feel like giving context every single time I wanna prompt a plan for a chapter or to write it. Summaries miss necessary info as well and the AI can miss context from even shorter stuff I send it. Since I’m trying to have something written for a specific series that definitely needs to be looked up, and locations aren’t really that simple honestly.
Could you recommend the best AI’s to use on mobile and on free tier to write ? At least one’s that are decently capable of following what I had in the style guide. I can still write or fix the rest. My fic averages around 6,500 words a chapter, so preferably I’d like a one capable of write that or at least a scene that’s more than 2000 words slightly. I don’t mind going scene by scene if at least it’s possible to do so as I can at best get two scenes done with Claude if I tried that method but not be able to tell it what’s wrong what it missed like for Deepseek, Deepseek can’t search at all which is an issue for me.
I only use open weight models on my own hardware, so I can’t give recommendations there from personal experience. If you want suggestions on open models though, there are uncensored Qwen and Gemma4 models that are pretty decent when run at higher quants. I use llmfan46’s
Qwen3.6-35B-A3B-uncensored-heretic-GGUFas my default and I’ve just downloadedgemma-4-26B-A4B-it-ultra-uncensored-heretic-GGUFthis past weekend to experiment with. (Gemma4 stock Q4_K_M has usually been annoyingly thick-headed compared to Qwen so I haven’t used Gemma4 as much, but the Q8 uncensored is pretty sharp, fast on my hardware, and unlike Qwen doesn’t get stuck in loops constantly, so I may switch over to that as my default… Need to evaluate more first though.)The sidebar gives a link-to-a-link to https://aihorde.net/ though – haven’t used it personally, but might be up your alley.
That writing guide is exactly why this isn’t working. You’re thinking about this whole thing wrong, which I pointed out in another post. This isn’t a single prompt problem. You need to separate all of the parts of your process into components.
Having the Tense, Mystery and Restraint, and how to use Punctuation in a single prompt is pure insanity.
You need to have about a dozen different agents, each focused on a specific task, with one coming up with the Mystery ideas, grabbing character background and information from another, and a third writing the main plot points out. Then, once you have the plot points, an editor goes through and identifies issues for correction, where it goes back to those original agents for revision. Once that’s been sorted, a new agent can write the initial story out into more detail, and another agent goes through and confirms tense and punctuation is consistent.
All of these agents should have multiple examples of what they’re expected to be outputting.
AI isn’t smart. You need to make it smart by building it properly.