Configure LLM settings for AI Chat

Overview

Large Language Model (LLM) Insights bring natural language intelligence to F5 Insight, enabling you to query your BIG-IP configurations and logs conversationally. Instead of manually searching through configurations or parsing log files, you can ask questions like “Why is pool member X marked down?” or “Show me all virtual IPs (VIPs) with SSL offloading enabled” and receive immediate, contextualized, clear answers.

LLM Insights also generates operational narratives and provides actionable recommendations for configuration best practices, security hardening, performance optimization, and troubleshooting next steps based on your environment’s data - all based on 30 years of F5 application delivery and security expertise. This robust operational intelligence helps teams streamline and improve operational workflows.

Benefits and key features

  • Reduced Mean Time to Resolution - Get answers to complex configuration and log questions in seconds rather than manually correlating data across multiple BIG-IP devices.
  • Simplified Management - Enable less experienced administrators to effectively troubleshoot and manage F5 infrastructure using natural language.
  • Actionable Recommendations - Receive context-aware suggestions for improving security posture, optimizing performance, and resolving issues based on your actual configurations and telemetry data.

Prerequisites

  • API Credentials - Obtain an API key or token from your chosen, currently supported LLM provider (OpenAI, Anthropic, or a local/self-hosted LLM).
  • Network Connectivity - Ensure F5 Insight can reach your LLM provider’s API endpoint (outbound HTTPS) and that any required firewall rules are in place for both inbound and outbound traffic.

Note

AI Chat tool-calling accuracy depends on the capability of the configured LLM. You can select a model from the available options for both OpenAI and Anthropic providers, or enter a custom model name. F5 Insight is validated against gpt-5.4 (OpenAI) and claude-sonnet-5-5 (Anthropic). For self-hosted local models, results might be inconsistent. Higher-parameter models generally perform better, and models with 120 billion or more parameters are recommended for reliable tool-calling support.

Before you begin

  • Sign in with an Administrator account or a role permitted to view and modify LLM settings.
  • Have your API key or token ready for your chosen provider.
  • For local LLM deployments, confirm that the model server is running and reachable from F5 Insight.

Configure the LLM provider and model

  1. Navigate to System Management > LLM Settings.
  2. Select Enable LLM Settings.
  3. Choose your LLM provider: OpenAI, Anthropic, or Local LLM.
  4. Enter the API Key or Token and, if needed, the API URL for your provider.
  5. Select a Model Name:
    • OpenAI and Anthropic: Select a model from the dropdown list. If your model isn’t listed, type its exact model ID in the search box and select Use custom model “<name>”.
    • Local LLM: Enter the model name your server exposes.
  6. Select Test Connection.
  7. Select Save Configuration.

Note

  • Model Name is required for all providers.
  • The field is pre-filled with a default for each provider (see the provider table below). Changing the provider resets the model to the new provider’s default.
  • Custom model IDs must exactly match one that your provider accepts. If the ID doesn’t match, Test Connection fails and shows the provider’s error message.
  • After you save, the configuration summary shows the selected model.

Provider configuration reference

Provider API URL Default model API key or token
OpenAI https://api.openai.com/v1 gpt-5.4 Obtain from platform.openai.com/api-keys
Anthropic https://api.anthropic.com claude-sonnet-5-5 Obtain from console.anthropic.com
Local LLM Varies by deployment (for example, http://acme.example.com:11434/v1 for Ollama) None — enter the model name Consult your local LLM documentation

Available models

The model dropdown offers the following models:

OpenAI:

  • gpt-5.4
  • gpt-6-astra
  • gpt-6.1-sol
  • gpt-6-luna
  • gpt-5.2
  • gpt-5

Anthropic:

  • claude-sonnet-5-5
  • claude-opus-5-5
  • claude-fable-5-1
  • claude-opus-4-8
  • claude-opus-4-6
  • claude-sonnet-4-6

To use any other model, type its ID in the search box. For current model IDs, see the OpenAI and Anthropic model documentation. Any OpenAI-compatible endpoint works with the Local LLM option.