Configuring Your AI Provider

YesBill supports three AI providers. You need an API key from at least one provider to use AI features.

Supported Providers

Provider	Models Available	Best For
OpenAI	GPT-4o, GPT-4o Mini, GPT-4.1	Speed and reliability
Anthropic	Claude 3.5 Sonnet, Claude 3 Haiku	Detailed analysis
Google	Gemini 2.0 Flash, Gemini 1.5 Pro	Budget-friendly
Ollama	Gemma 4, Llama 3.2, Qwen3 Coder, and any locally installed model	Free, fully local — no API key

Getting an API Key

OpenAI

1. Visit platform.openai.com
2. Sign up or log in
3. Go to API Keys → Create new secret key
4. Copy the key (starts with sk-...)

Anthropic

1. Visit console.anthropic.com
2. Sign up or log in
3. Go to API Keys → Create Key
4. Copy the key

Google Gemini

1. Visit aistudio.google.com
2. Click Get API Key
3. Copy the key

Adding Your Key to YesBill

Go to Settings → AI Configuration to see the full configuration panel. From here you can manage providers, models, reasoning settings, and AI insights.

Step 1 — Choose a Provider

Select your preferred AI provider by clicking its tab (OpenAI, Anthropic, or Google). Then paste your API key into the field provided and click Validate to verify it works.

1. Go to Settings → AI Configuration
2. Select your provider tab
3. Paste your API key
4. Click Validate to test the key
5. Click Save

Step 2 — Select a Default Model

After adding your API key, choose the default model you want YesBill to use for all AI features. The available models update automatically based on your validated provider.

tip

Start with OpenAI GPT-4o Mini or Google Gemini 2.0 Flash for low-cost usage while you're getting started.

Step 3 — Set Default Reasoning Effort

For models that support extended thinking (like Claude 3.7 Sonnet or o3-mini), you can set the default reasoning effort:

Level	Behaviour	Best For
Low	Fast, minimal reasoning	Quick bill lookups
Medium	Balanced speed and depth	General Ask AI use
High	Deep reasoning, slower	Complex spending analysis

Choose Medium as a sensible default for most households.

Step 4 — Enable or Disable AI Insights

The AI Insights toggle controls whether YesBill automatically surfaces AI-generated insights on your Dashboard and Bills pages (e.g. "Your milk spend increased 12% this month"). Toggle it off if you prefer a cleaner interface or want to reduce AI API usage.

Multiple Providers

You can add keys for all three providers and switch between them in Ask AI using the model selector. Each provider's key is stored separately and securely.

Security

Your API keys are stored only in your browser session and sent directly to the AI provider. YesBill never stores, logs, or shares your API keys. Your keys remain private and under your control at all times.

Removing a Key

1. Go to Settings → AI Configuration
2. Click Remove next to the provider
3. Confirm the removal

After removing all keys, AI features will be disabled until you add a new key.

---

Ollama (Local Models)

Ollama lets you run AI models entirely on your own machine — completely free, no API key, no usage costs.

Why use Ollama?

• Free — no per-token charges ever
• Private — your data never leaves your machine
• Offline capable — works without internet once set up
• Any model — pull any model from the Ollama library

How it works

YesBill connects to your local Ollama instance through a Cloudflare Tunnel, which creates a secure public URL that the cloud backend can reach.

Setting up Ollama

1. Go to Settings → AI Configuration
2. Click the Ollama provider tile
3. Paste your Cloudflare Tunnel URL in the Ollama Base URL field
4. Click Fetch Models to load your installed models
5. Select a model and click Save Changes

info

Ollama requires a one-time local setup. See the full guide: Ollama Setup →

Recommended models for YesBill

Model	Size	Best For
gemma4:latest	~5 GB	General use (recommended)
llama3.2	~2 GB	Fast responses
qwen3-coder:latest	~4 GB	Code and bill analysis
gemma3:latest	~3 GB	Balanced

Ollama vs cloud providers

	Ollama	OpenAI / Anthropic / Google
Cost	Free	Pay per token
Privacy	Fully local	Sent to provider servers
Speed	Depends on your GPU	Fast cloud inference
Setup	Requires local install	API key only
Offline	Yes (after setup)	No