What multi-provider means

A multi-provider coding agent is not locked to one model vendor. It can route requests to different providers or models depending on the task, the phase, cost, latency or privacy requirements. The agent workflow stays the same; the model backend can change.

Bytos is a multi-provider coding agent and an autonomous coding agent. The two categories are connected: a multi-step task may use one model for planning and another for implementation or review.

Verified providers

The current product supports OpenRouter, Anthropic, OpenAI, Google Gemini, AWS Bedrock, Azure, GCP Vertex, Cerebras, Groq, any OpenAI-compatible endpoint, and local models through Ollama or LM Studio. This list is verified against the product README and provider configuration.

Provider availability and model names can change. Check the current provider documentation and the Marketplace listing before relying on a specific model.

Per-task and per-phase models

You can choose a provider and model for the task. In sequential Auto Phase, each phase can have its own provider, model, prompt, attachments and MCP allowlist. A planning phase might use a stronger reasoning model, an implementation phase a faster coding model, and a review phase a different provider.

Auto Phase is sequential, not a parallel multi-agent system. The phases run in order, and a run can be stopped.

Keys and local models

Provider API keys are stored in VS Code SecretStorage. The webview receives a length-based mask, not the raw key. With your own provider, requests go directly to that provider; optional third-party compatibility providers use their own hosted API. Local models through Ollama or LM Studio keep inference on your machine, subject to the tools you approve.

BYOK is a technical capability here, not the category. The category is the multi-provider architecture: one agent, multiple model backends.

Current platform

Bytos currently runs as a VS Code extension. JetBrains, Cursor, Windsurf, CLI and a standalone IDE are roadmap targets, not current integrations. Do not use installation instructions for unsupported environments.

Install Bytos for VS Code

Bytos is currently a VS Code extension. Install it from the Marketplace, connect your own provider and start with an approval-first workflow.

FAQ

Is Bytos locked to one provider?

No. It supports multiple providers and models, including local models. You choose what each task or phase uses.

Does multi-provider mean parallel agents?

No. Multi-provider refers to model backends. Auto Phase is sequential; Bytos does not claim a parallel multi-agent system.

Are provider keys stored in the browser?

No. Provider keys are stored in VS Code SecretStorage. The webview receives a masked value.