AI capabilities
What does bring your own model mean?
Pointing the tool at your own model deployment or LLM API key instead of the vendor's hosted inference.
Bring-your-own-model routes inference through infrastructure you control, such as your Azure OpenAI deployment, your Bedrock account or a self-hosted open-weights model. Your code goes to your tenancy rather than the vendor's. Some vendors call this BYOK, meaning your own LLM API key, which is a different thing from customer-managed encryption keys.
This is usually a compliance requirement rather than a preference: it keeps inference inside an existing data-processing agreement and an audited boundary. Where it is offered it is often limited to enterprise plans, and worth confirming exactly which features still work when enabled.
What to look for
- Which providers and models are supported
- Whether all AI features work with your model, or only a subset
- Who bears inference cost and how that changes the total price
- Whether prompts and responses still transit the vendor's systems
Best 11 code quality, security and code review tools that let you bring your own model
Every tool in the directory whose documentation covers this, 10 fully and 1 partially. Full support is shown first. Shown in a random order that changes on every visit, so no tool gets an advantage from its position.
Support is not the same as parity, since some implementations are narrower in scope, gated to a higher plan tier, or maintained only for existing customers. The note under each tool is what its own documentation describes.
AI Coding Agent
ChatGPT / CodexRuns against OpenAI API, Amazon Bedrock, OpenRouter or Fireworks; review_model overrides the review model.
AI Coding Agent
Claude CodeBring your own key via Console, Bedrock, Vertex, Foundry or LLM gateways; Claude models only.
PR Review Tool
CodeRabbitBring-your-own-LLM provider available only in self-hosted Enterprise deployments
Code Quality Platform
CodeSceneUser controls LLM provider and model for the PR Refactoring Agent; CodeHealth MCP is model-agnostic
Code Quality Platform
DeepSourceEnterprise Server v5.0.0+ BYOK routes inference through Vertex AI, Bedrock, Azure OpenAI, OpenAI, Google AI, or Anthropic.
AI Coding Agent
Gemini Code AssistBuyer can supply their own Gemini API key for the CLI; Gemini models only.
PR Review Tool
GreptileAny OpenAI-compatible or custom LLM (Anthropic, OpenAI, Azure OpenAI, Bedrock); self-hosted only
PR Review Tool
QodoOn-prem installs connect to customer OpenAI, Anthropic, Vertex AI or AWS Bedrock endpoints
Code Security Platform
SemgrepBYO key for OpenAI, Bedrock, Azure OpenAI, Gemini, xAI; Enterprise; forfeits fine-tuned models and ZDR.
Code Quality Platform
SonarQubeAzure OpenAI/AWS Bedrock/self-hosted gateway BYOK on Server Enterprise/Data Center only; Cloud has no BYOK option
Code Security Platform
Checkmarx OnePartialAzure AI selectable at tenant level and Developer Assist switchable between GitHub Copilot and Claude; no customer-key BYOK.
Related terms
FAQ
What does bring your own model mean?
Pointing the tool at your own model deployment or LLM API key instead of the vendor's hosted inference.
How many tools in the directory let you bring your own model?
11 of the 20 tools in this directory let you bring your own model, 10 fully and 1 partially:
Support is not the same as parity, since some implementations are narrower in scope, gated to a higher plan tier, or maintained only for existing customers. The note under each tool is what its own documentation describes.
What should you check when comparing tools that let you bring your own model?
Four things:
- Which providers and models are supported
- Whether all AI features work with your model, or only a subset
- Who bears inference cost and how that changes the total price
- Whether prompts and responses still transit the vendor's systems