{% extends "base.html" %} {% block title %}abogen · Settings{% endblock %} {% block content %} Application Settings Settings apply to new jobs you queue from the dashboard. {% if saved %} Settings saved successfully. {% endif %} Narration Audio & Delivery Subtitles & Text Performance LLM Text Normalization Narration Defaults Narrator Voice {% for voice in options.voices %} {{ voice }} {% endfor %} Used whenever “Standard voice” is selected for a new job. Chunk Granularity {% for option in options.chunk_levels %} {{ option.label }} {% endfor %} Speaker Analysis Minimum Mentions Speakers detected fewer times fall back to the narrator voice. Speaker Pronunciation Preview Include {{ '{{name}}' }} where the speaker name should be inserted. Automatically detect entities for new jobs Disable if you prefer to skip entity extraction in the job wizard. Randomizer Languages {% set selected_languages = settings.speaker_random_languages or [] %} {% for code, label in options.languages.items() %} {{ label }} ({{ code|upper }}) {% endfor %} Limits random voice selection for speakers marked as random. Leave empty to allow any language. Audio & Delivery Audio Format {% for fmt in options.output_formats %} {{ fmt }} {% endfor %} Save Location {% for location in save_locations %} {{ location.label }} {% endfor %} Default output: {{ default_output_dir }} Separate Chapter Format {% for fmt in options.separate_formats %} {{ fmt|upper }} {% endfor %} Save Each Chapter Separately Also Create Merged Audiobook Save as Project With Metadata Silence Between Chapters (Seconds) Pause After Chapter Titles (Seconds) Inserted between the spoken chapter title and the chapter content. Set to 0 to disable. Add "Chapter" before numeric chapter titles Ensures the spoken chapter heading starts with "Chapter" when source titles begin with only a number or numeral. Subtitles & Text Subtitle File Format {% for value, text in options.subtitle_formats %} {{ text }} {% endfor %} Max Words Per Subtitle Entry Replace Single Newlines Generate EPUB 3 (experimental) Performance Use GPU (when available) Endpoint Base URL Point to an OpenAI-compatible endpoint such as Ollama or a proxy. API Key Leave blank or use ollama for local servers that do not require keys. Default Model {% if settings.llm_model %} {{ settings.llm_model }} {% else %} Select a model {% endif %} Timeout (seconds) Refresh models Normalization Prompt Prompt Template {{ settings.llm_prompt }} Use placeholders like {{ '{{sentence}}' }} and {{ '{{paragraph}}' }} to inject content. Context Mode {% for option in llm_context_options %} {{ option.label }} {% endfor %} Try the prompt I've been waiting all day. Preview Normalization Rules Convert grouped numbers to words Expand titles and suffixes (Dr., St., Jr., …) Ensure sentences end with terminal punctuation Add phoneme hints for possessives Apostrophe strategy {% for option in apostrophe_modes %} {{ option.label }} {% endfor %} {% if settings.normalization_apostrophe_mode == 'llm' and not llm_ready %} Configure the LLM connection before using it for audiobook runs. {% endif %} Sample & Preview Sample {% for key, text in normalization_samples.items() %} {{ key|capitalize }} {% endfor %} Voice {% for voice in options.voices %} {{ voice }} {% endfor %} Sample text {{ normalization_samples['apostrophes'] }} Preview with current settings Save Settings {% endblock %} {% block scripts %} {{ super() }} {% endblock %}
Settings apply to new jobs you queue from the dashboard.
Used whenever “Standard voice” is selected for a new job.
Speakers detected fewer times fall back to the narrator voice.
Include {{ '{{name}}' }} where the speaker name should be inserted.
{{ '{{name}}' }}
Disable if you prefer to skip entity extraction in the job wizard.
Limits random voice selection for speakers marked as random. Leave empty to allow any language.
Default output: {{ default_output_dir }}
{{ default_output_dir }}
Inserted between the spoken chapter title and the chapter content. Set to 0 to disable.
Ensures the spoken chapter heading starts with "Chapter" when source titles begin with only a number or numeral.
Point to an OpenAI-compatible endpoint such as Ollama or a proxy.
Leave blank or use ollama for local servers that do not require keys.
ollama
Use placeholders like {{ '{{sentence}}' }} and {{ '{{paragraph}}' }} to inject content.
{{ '{{sentence}}' }}
{{ '{{paragraph}}' }}
Configure the LLM connection before using it for audiobook runs.