Optional AI — Always Your Choice
Fetch & Flow Lore Studio works fully without AI.
All core tools for organising lore, characters, locations, and story arcs operate entirely offline.
When you want AI assistance, you have two flexible options:
- Run AI locally using Ollama with open models such as Gemma, Qwen, and Llama
- Connect cloud models using your own API key from OpenAI, Anthropic, or Google Gemini. Which also includes Elevenlabs API keys for realistic voice generation .
This gives creators full control over privacy, costs, and workflow.
Running AI Locally
Fetch & Flow Lore Studio can run AI models directly on your computer using Ollama.
Ollama allows modern language models to run locally without sending your writing to cloud AI providers. This enables AI-assisted writing, brainstorming, and lore generation directly on your own machine.
Running AI locally provides several advantages:
-
Greater privacy (your writing stays on your computer)
-
No API usage costs
-
The ability to work offline
Fetch & Flow Lore Studio supports a curated set of Ollama models designed for different hardware capabilities.
You can learn more about Ollama by visiting their website: Ollama
Supported Local Models
The models below are organised by hardware capability.
Performance will depend on your computer’s RAM, CPU, and GPU.
Fast (Lightweight Models)
These models are designed to run on most modern laptops and desktops.
Recommended for quick brainstorming and lighter writing assistance.
Recommended hardware:
8-16 GB RAM
Supported models:
Balanced
These models provide stronger responses while still running on many consumer computers.
Recommended for everyday worldbuilding and writing assistance.
Recommended hardware:
16 GB RAM
Supported models:
-
Qwen 2.5 3B
-
Llama 3.2 3B
-
Gemma 2 2B
-
Gemma 3 4B
Best (Strong Local Performance)
These models offer stronger reasoning and writing quality but require more memory.
Recommended hardware:
16–32 GB RAM
Supported models:
-
Qwen 2.5 7B
-
Gemma 2 9B
-
Gemma 3 12B
-
Mistral Nemo
Stronger Machines
These models provide significantly stronger results but require high-end desktop hardware.
Recommended hardware:
32+ GB RAM
Supported models:
-
Gemma 3 27B
-
Gemma 4 26B
Very High-End Systems
These models require powerful hardware and large amounts of memory.
They are intended for advanced users running dedicated AI systems.
Recommended hardware:
64+ GB RAM or GPU-accelerated systems
Supported models:
-
Gemma 2 27B
-
Gemma 4 31B
When Should I Use Local AI?
Local AI works well for:
-
brainstorming ideas
-
expanding scenes
-
drafting lore entries
-
experimenting with creative prompts
If you prefer the most powerful models available, you can still connect cloud providers such as OpenAI, Anthropic, or Google using your own API key.
How to Set Up Local AI
-
Install Ollama from
https://ollama.com -
Download a supported model using the Ollama command line:
Example commands:
ollama run gemma3:4b
ollama run qwen2.5:3b
ollama run mistral-nemo
-
Open Fetch & Flow Lore Studio
-
Go to AI Settings
-
Select Local AI (Ollama)
-
Choose one of the supported models
Hardware Considerations
Local AI performance varies depending on your system.
General recommendations:
-
8–16 GB RAM: lightweight models
-
16–32 GB RAM: balanced and stronger models
-
32+ GB RAM: large local models
Smaller models will generate responses faster and are recommended for most users.
Using Cloud AI Providers
(API Keys)
In addition to running AI locally, Fetch & Flow Lore Studio also supports connecting to cloud-based AI providers using your own API keys.
This allows you to access hosted AI models from major providers if you prefer cloud-based processing or want to experiment with different model capabilities.
When using cloud AI, only the text you choose to send is transmitted to the selected provider using your own API key.
Fetch & Flow does not issue API keys or manage AI billing.
All usage costs are billed directly by the provider associated with your key.
Supported Cloud AI Providers
Fetch & Flow Lore Studio currently supports API keys from the following providers:
-
OpenAI
-
Anthropic
-
Google Gemini
-
ElevenLabs (Voice Studio)
Each provider offers different models and pricing options.
Additional providers may be added over time.
Why Bring Your Own API Key?
Fetch & Flow uses a Bring Your Own Key (BYOK) approach.
This ensures:
-
You control which AI provider you use
-
You manage your own usage and billing
-
Your writing is not stored by Fetch & Flow
-
There are no hidden subscriptions or usage limits imposed by the app
Fetch & Flow simply acts as a client that connects your writing tools to the provider you choose.
Important Notes About API Keys
-
Never share your API key publicly
-
Your API key is stored locally on your machine
-
You can remove or replace your key at any time
-
All AI usage costs are billed directly by the provider
What Happens to Your Writing?
Your writing always remains under your control.
-
Your text stays on your computer
-
AI requests are only sent when you choose to use AI features
-
Fetch & Flow does not store your writing or API keys on external servers
Your writing is never uploaded automatically in the background.
Getting an API Key
Below are guides for obtaining API keys from supported providers.
You only need to follow the steps for the provider you wish to use.
How to Get an OpenAI API Key
Fetch & Flow Lore Studio does not include AI access by default.
To use optional AI features with OpenAI models, you will need to create an API key from your OpenAI account.
Step-by-step
-
Visit the OpenAI API key page:
https://platform.openai.com/api-keys -
Sign in or create an OpenAI account.
-
Click Create new secret key.
-
Copy the generated API key.
-
Paste the key into Fetch & Flow Lore Studio.
Your API key can be revoked or replaced at any time from your OpenAI dashboard.
Useful Youtube video that helps explain generating your own OpenAI Key
Official documentation:
https://platform.openai.com/docs
How OpenAI Handles Your Data
When using AI features in Fetch & Flow Lore Studio, only the text you choose to send is transmitted to OpenAI through your API key.
More information:
https://platform.openai.com/docs/guides/your-data
How to Get an Anthropic API Key
Fetch & Flow Lore Studio also supports models from Anthropic.
To use these models, you will need an API key from Anthropic.
Step-by-step
-
Visit the Anthropic API console:
https://console.anthropic.com/settings/keys -
Sign in or create an Anthropic account.
-
Click Create Key.
-
Copy the generated key.
-
Paste the key into Fetch & Flow Lore Studio.
Useful Youtube video that helps explain generating your own Claude API Key
You can revoke or rotate your key at any time from the Anthropic dashboard.
How Anthropic Handles Your Data
When using Anthropic models, the text you send through Lore Studio is transmitted directly to Anthropic through your API key.
Privacy policy:
https://www.anthropic.com/legal/privacy
How to Get a Google Gemini API Key
Fetch & Flow Lore Studio also supports Google Gemini models through Google AI Studio.
Step-by-step
-
Visit Google AI Studio:
https://aistudio.google.com -
Sign in with your Google account.
-
Open the API Keys section:
https://aistudio.google.com/app/apikey -
Create a new API key.
-
Copy the key and paste it into Fetch & Flow Lore Studio.
Useful Youtube video that helps explain generating your own Gemini API Key
You can manage or revoke your keys through Google AI Studio.
How Google Handles Your Data
When using Gemini models through Fetch & Flow, the text you choose to send is transmitted to Google's Gemini API.
More information:
https://ai.google.dev/gemini-api/terms
https://ai.google.dev/gemini-api/docs/privacy
Using Voice Studio with ElevenLabs
Voice Studio features use ElevenLabs text-to-speech.
To enable these features you must provide your own ElevenLabs API key.
Step-by-step
-
Create an account at
https://elevenlabs.io -
Open Profile Settings
-
Locate your API Key
-
Copy the key
-
Paste the key into Fetch & Flow Lore Studio
Voice generation costs are billed directly by ElevenLabs under your own account.