AI Model Providers

Knowi supports multiple AI model providers for powering features like InstantSights, AI Recommendations, AI Chat, Document AI, and more. By default, Knowi uses its built-in AI service. You can optionally configure external providers for enhanced capabilities.

Supported Providers

Configuring AI Providers

To configure an external AI provider:

1. Navigate to User Settings from the bottom-left sidebar.
2. Open the AI Settings tab.
3. Under the provider section, click CHANGE next to the provider you want to configure (OpenAI, Anthropic, Google, or Azure AI).
4. Enter your API key in the modal and click Save. For Azure AI, enter your endpoint URL as well
5. Once a key is saved, the corresponding models will appear in the Default AI Model Selection dropdowns for each AI feature.

Per-Feature Model Selection

Knowi allows you to select different AI models for different features. Under Default AI Model Selection in AI Settings, you can assign a provider for each of the following:

• NLP Search - Powers natural language queries across datasets.
• AI Chat - Powers the long-form AI chat experience.
• Insights - Powers InstantSights summaries.
• Recommendation Widget - Powers AI-generated recommendations.
• Documents - Powers Document AI question answering.
• Dashboards - Powers AI Dashboard generation.
• Agents - Powers Agentic AI assistants.

For example, you could use Knowi AI for NLP Search (to keep data private) while using Claude Sonnet 5 for AI Recommendations (for more advanced analysis).

Anthropic (Claude) Setup

1. Obtain an API key from Anthropic Console.
2. In Knowi AI Settings, click CHANGE next to the Anthropic (Claude) row.
3. Enter your API key and save.
4. Select a Claude model from the feature dropdowns.

Available Claude Models:

Reasoning Effort

Claude models can spend more or less time reasoning before answering. Set Reasoning Effort in the Anthropic (Claude) section of AI Settings to control that tradeoff across every feature using a Claude model:

• Low - Fastest and cheapest. Best for short, well-scoped tasks.
• Medium - Reduces token usage where the task does not need deep reasoning.
• High - The default when no effort is set. A good balance for most analytical work.
• Extra High - Recommended for the hardest analysis and agent tasks.
• Max - Highest capability, highest cost. Best when correctness matters more than speed.

Leave the setting empty to use each model's default. If you select an effort level a model does not support, Knowi automatically steps down to the closest supported level rather than failing the request ? Extra High becomes High on Claude Sonnet 4.6 and Claude Opus 4.6, and effort is skipped entirely on Claude Haiku 4.5.

Task Budget

Task Budget caps how many tokens a single Claude request may spend, including reasoning. Set it in tokens in the Anthropic (Claude) section of AI Settings. The minimum accepted value is 20,000 tokens; values below that, or non-numeric entries, are ignored and the request runs without a budget.

Google (Gemini) Setup

1. Obtain an API key from Google AI Studio.
2. In Knowi AI Settings, click CHANGE next to the Google (Gemini) row.
3. Enter your API key and save.
4. Select a Gemini model from the feature dropdowns.

Available Gemini Models:

The Gemini 2.5 models remain selectable so existing configurations keep working, but new setups should use the 3.x models. Google has already restricted some 2.5 models to accounts that used them previously; if a 2.5 model returns an error saying it is no longer available to new users, switch to the 3.x model listed above as its replacement.

Gemini models can be used for every AI feature, including Document AI. On-premises deployments need a Knowi AI service build that includes Gemini support before Document AI can answer over a Gemini model; contact your Knowi administrator if a Gemini selection works elsewhere but not in Document AI.

Note: Gemini models are subject to Google's own per-model availability and quota. On the Google free tier some models return no output at all rather than an error. When a request to an external provider fails, Knowi now surfaces the provider's own error message rather than a generic one, so the reason - invalid key, model unavailable, quota exhausted - is visible directly in the feature you were using. Internal Knowi AI errors continue to show the generic message, since they are not actionable by the user.

OpenAI Setup

1. Obtain an API key from OpenAI Platform.
2. In Knowi AI Settings, click CHANGE next to the OpenAI row.
3. Enter your API key and save.
4. Select an OpenAI model from the feature dropdowns.

Available OpenAI Models:

OpenAI Effort

OpenAI reasoning models can spend more or less time thinking before answering. Set OpenAI Effort in the OpenAI section of AI Settings to control that tradeoff across every feature using an OpenAI model:

• Low - Fastest and cheapest. Best for straightforward, high-volume work.
• Medium - Balanced reasoning depth against speed and cost.
• High - Deepest reasoning. Best for complex analysis, at the cost of more tokens and more time.

Deprecated OpenAI Models

o3 and o3-mini are scheduled for retirement by OpenAI. o3-mini retires on October 23, 2026, and the o3 snapshot on December 11, 2026. OpenAI recommends GPT-5.6 Sol as the replacement for both. They remain selectable so existing configurations keep working, but new setups should use a GPT-5.6 model. For the current schedule, see OpenAI's deprecations page.

When OpenAI fully retires a model, it is removed from the model lists in AI Settings. Any feature still pointing at the removed model falls back to Knowi AI until you choose a new one, so review your Default AI Model Selection after a retirement.

Azure AI Setup

1. In the Azure portal, open your Azure OpenAI resource and go to Keys and Endpoint to copy your API key and endpoint URL (for example, https://myresource.openai.azure.com).
2. Deploy the models you want to use in your Azure resource, using their standard names: gpt-5.5, gpt-5.4-mini, gpt-5, gpt-4o, gpt-4.1.
3. In Knowi AI Settings, click CHANGE next to the Azure AI row.
4. Enter your API key and endpoint URL, then save.
5. Select an Azure model from the feature dropdowns.

Note: Unlike other providers, Azure AI requires both an API key and an endpoint URL. Models must be deployed in your Azure resource before Knowi can use them; if a model is not deployed, requests will fail with a deployment not found error.