FAQ & Troubleshooting¶

Common Issues¶

"Connection failed" or "API error"¶

Check your API key in Settings → Providers. No extra spaces.
Test the connection — click "Test Connection" for specific error details.
Check your network — if you're behind a firewall or VPN, the API endpoint may be blocked.
Key expired? Check your provider's dashboard for remaining credits.

Generation produces poor or repetitive results¶

Improve your premise — be specific about characters, setting, tone, and conflict.
Add creative direction — steer tone with "noir atmosphere, dry humor."
Check character cards — run Quality Inspection (🛡) to find gaps.
Try a different writing style — each dramatically changes the AI's approach.
Try a better model — higher-quality models produce significantly better fiction. See Choosing a Model.

Lorebook entries aren't activating¶

Check keywords — they must match text in the current scene context (character names, outline, summaries).
Check the entry is enabled — disabled entries won't trigger.
Check scope — world entries need the story to use that world; character entries need the character assigned to the story.

Characters seem inconsistent¶

Check Story Context → World — the AI may have extracted incorrect state. Click Manage events to delete wrong entries.
Check Story Context → Events — see what the AI "remembers." Missing facts may have been missed during extraction.
Add critical facts as lorebook entries — keyword-triggered entries are always included when matched; memories are probabilistic.

Local models not connecting (Ollama, LM Studio)¶

Browsers block cross-origin requests to localhost. Enable CORS:

LM Studio — Developer tab → enable "Allow requests from any origin (CORS)"
Ollama — Set OLLAMA_ORIGINS=* before starting

Data not persisting¶

Web storage is browser-local. Your data won't sync across browsers or survive clearing browser data. Export your stories regularly.

Google Gemini 400 error¶

If you see "Multiple authentication credentials," make sure you're using the Google Gemini provider preset (not Custom or OpenAI) — it uses the correct authentication header.

FAQ¶

How much does it cost?¶

Storywright is free to try — demo mode lets you write without an API key. For unlimited use, bring your own API key. A typical 10-scene story costs $0.50–$2.00 with mid-range models. See Choosing a Model for budget tiers.

Is my data private?¶

Your stories are stored in your browser. AI prompts are sent directly to your API provider — there's no Storywright server in between. Your API key is stored locally and only sent to the endpoint you configured.

Can I use local/offline AI?¶

Yes. Set up an Ollama or LM Studio provider in Settings. Local models are free but need sufficient hardware — and may produce lower-quality fiction than cloud models.

What models work best?¶

See Choosing a Model for recommendations by provider and budget.

Can I export my stories?¶

Yes — Markdown, plain text, or copy to clipboard. Find the export menu in the toolbar.

Can I import from other tools?¶

Character cards — PNG or JSON files (SillyTavern-compatible)
Lorebooks — JSON files
Chat transcripts — JSONL or plain text (see Writing Your Story)

What's demo mode?¶

Demo mode uses a built-in API key so you can start writing immediately without creating an account or entering payment info. It has a word limit and pre-selected models. To unlock full access, add your own API key in Settings → Providers — demo mode lifts automatically.

How do I back up my data?¶

On web, export your stories regularly. Browser local storage doesn't persist across browsers or survive data clearing.