AWS Bedrock Chat Provider for VS Code

Use AWS Bedrock models directly in VS Code chat via:

• Mantle (OpenAI-compatible API) for the OSS/openai-style model catalog
• Native Bedrock (Converse API) for the full Bedrock foundation model catalog

Features

• Mantle + Native: Use Mantle models and native Bedrock foundation models
• Dynamic Model Discovery: Mantle models are fetched from Mantle's Models API; native models are listed from AWS Bedrock
• Streaming Responses: Real-time chat with streaming support
• Tool Calling: Function calling support for capable models
• Multi-Region: Support for 12 AWS regions
• OpenAI Compatible (Mantle): Uses familiar OpenAI SDK patterns via Mantle
• Converse API (Native): Uses the unified Bedrock conversation API

Available Models

OpenAI

• gpt-oss-20b, gpt-oss-120b
• Safeguard variants: gpt-oss-safeguard-20b/120b

Google

• Gemma 3: 4b, 12b, 27b variants

Mistral

• magistral-small-2509
• mistral-large-3-675b-instruct
• Ministral: 3b, 8b, 14b variants
• Voxtral: mini-3b, small-24b variants

Qwen

• General: qwen3-32b, qwen3-235b, qwen3-next-80b
• Vision: qwen3-vl-235b (multimodal)
• Coding: qwen3-coder-30b/480b

DeepSeek

• v3.1

Nvidia

• nemotron-nano-9b-v2, nemotron-nano-12b-v2

Others

• MoonshotAI: kimi-k2-thinking
• Minimax: minimax-m2
• ZAI: glm-4.6

Prerequisites

You can use either backend (or both):

1. Mantle (optional):
   • Option A: AWS Bedrock API Key from the AWS Bedrock Console (simpler)
   • Option B: AWS credentials/profile (better for existing AWS setups)
2. Native Bedrock (optional): AWS credentials available to VS Code (env vars, ~/.aws/credentials, SSO, etc). You can also set aws-bedrock.awsProfile.
3. VS Code: Version 1.104.0 or later

Installation

From VS Code Marketplace

1. Open VS Code
2. Go to Extensions (Cmd+Shift+X)
3. Search for "Bedrock LLMs for VS Code Chat"
4. Click Install

From Source

1. Clone this repository:

   git clone https://github.com/easytocloud/bedrock-vscode-chat.git
   cd bedrock-vscode-chat
   

2. Install dependencies:

   npm install
   

3. Compile the extension:

   npm run compile
   

4. Press F5 to open a new VS Code window with the extension loaded

Setup

1. Configure Mantle Authentication (Optional)

Mantle supports two authentication methods:

Option A: API Key (Simpler)

Via Command Palette:

1. Open Command Palette (Cmd+Shift+P / Ctrl+Shift+P)
2. Run: Manage AWS Bedrock
3. Select "Enter API Key (Mantle)"
4. Paste your API key from AWS Bedrock Console

On First Use:

• The extension will prompt for your API key when you first try to use a Mantle model
• Your key is stored securely in VS Code's SecretStorage

Option B: AWS Credentials (Better for existing AWS setups)

1. Open Command Palette
2. Run: Manage AWS Bedrock
3. Select "Configure Mantle Authentication"
4. Choose "AWS Credentials"
5. Optionally set a specific profile via "Set AWS Profile (Mantle)"

This method uses AWS Signature V4 authentication with your existing AWS credentials.

2. Configure Native AWS Profile (Optional)

If you're using native Bedrock models and want a specific named profile:

1. Run: Manage AWS Bedrock
2. Select "Set AWS Profile (Native)"
3. Enter a profile name (or leave blank to use the default credential chain)

3. Select Region (Optional)

Default region is us-east-1. To change:

1. Open Command Palette
2. Run: Manage AWS Bedrock
3. Select "Change Region"
4. Choose your preferred AWS region

Or set in Settings:

{
  "aws-bedrock.region": "us-west-2",
  "aws-bedrock.mantleAuthMethod": "awsCredentials",  // or "apiKey"
  "aws-bedrock.mantleAwsProfile": "my-profile",      // optional
  "aws-bedrock.awsProfile": "my-profile"             // for native Bedrock
}

4. Configure Model Visibility (Optional)

Show/hide specialized models (like safeguard variants):

{
  "aws-bedrock.showAllModels": true  // default: true
}

Usage

Using in Chat

1. Open VS Code Chat (Cmd+Shift+I / Ctrl+Shift+I)
2. Click the model picker (top of chat panel)
3. Select an AWS Bedrock model (e.g., "OpenAI GPT OSS 120B")
4. Start chatting!

Using with Copilot Chat

1. In any editor, use @workspace or other chat participants
2. The model picker will include Bedrock models
3. Select a Bedrock model for your conversation

Example Chat

You: What are the key features of Rust's ownership system?

Assistant (via Bedrock): [Streams response in real-time...]

Configuration

Settings

Setting	Type	Default	Description
aws-bedrock.region	string	us-east-1	AWS region for Bedrock (Mantle + native)
aws-bedrock.enableMantle	boolean	true	Enable Mantle (OpenAI-compatible) models
aws-bedrock.enableNative	boolean	true	Enable native Bedrock models via Converse API
aws-bedrock.mantleAuthMethod	string	apiKey	Mantle auth method: apiKey or awsCredentials
aws-bedrock.mantleAwsProfile	string	empty	Optional AWS profile for Mantle when using credentials
aws-bedrock.awsProfile	string	empty	Optional AWS profile for native Bedrock
aws-bedrock.showAllModels	boolean	true	Show all models including specialized variants
aws-bedrock.debugLogging	boolean	false	Enable verbose debug logging
aws-bedrock.sendTools	boolean	true	Send tool definitions to the model
aws-bedrock.emitPlaceholders	boolean	true	Emit placeholder text while waiting
aws-bedrock.modelMetadataSource	string	litellm	Metadata source for token/capability info
aws-bedrock.modelMetadataUrl	string	default URL	External metadata registry URL
aws-bedrock.modelMetadataCacheHours	number	24	Cache duration for external metadata

Supported Regions

• us-east-1 (N. Virginia) - Default
• us-east-2 (Ohio)
• us-west-2 (Oregon)
• eu-west-1 (Ireland)
• eu-west-2 (London)
• eu-central-1 (Frankfurt)
• eu-north-1 (Stockholm)
• eu-south-1 (Milan)
• ap-south-1 (Mumbai)
• ap-northeast-1 (Tokyo)
• ap-southeast-3 (Jakarta)
• sa-east-1 (São Paulo)

Commands

Command	Description
Manage AWS Bedrock	Configure Mantle auth, native AWS profile, region, and settings
Clear AWS Bedrock API Key (Mantle)	Remove stored Mantle API key
Show AWS Bedrock Logs	Open the extension output channel

Architecture

This extension implements VS Code's LanguageModelChatProvider interface using AWS Bedrock's Mantle API, which provides OpenAI-compatible endpoints.

Key Components

• BedrockMantleProvider: Main provider implementing VSCode's chat interface
• Dynamic Model Discovery: Fetches available models from Mantle's Models API
• Streaming Support: Processes SSE (Server-Sent Events) for real-time responses
• Tool Calling: Buffers and parses streaming tool calls for function calling support

Endpoint Format

https://bedrock-mantle.<region>.api.aws/v1

Model Capabilities

Tool Calling Support

Models with function calling capabilities:

• gpt-oss-120b
• mistral-large-3-675b-instruct
• magistral-small-2509
• deepseek.v3.1
• qwen3-235b and larger models
• qwen3-vl-235b (vision + tools)

Vision Support

Models with multimodal (image) input:

• Mantle models: based on model naming and API behavior
• Native Bedrock models: based on Bedrock's reported input modalities

Notes on Capability Metadata

• Token limits + initial capabilities: The extension can optionally use an external model metadata registry (default: Litellm's public JSON) to populate maxInputTokens, maxOutputTokens, and initial tool/vision flags. Configure via aws-bedrock.modelMetadataSource, aws-bedrock.modelMetadataUrl, and aws-bedrock.modelMetadataCacheHours.
• Native Bedrock models: vision is derived from ListFoundationModels input modalities (reliable). Tool support is still verified on-demand by attempting a tool-enabled request and caching whether the model accepts tool config (this overrides external metadata if the runtime behavior differs).
• Mantle models: Mantle's /v1/models does not include tool/vision/token metadata, so the extension uses external metadata when enabled, plus runtime probing (tools) as a safety net.

Code Specialization

Models optimized for coding:

• qwen3-coder-30b-a3b-instruct
• qwen3-coder-480b-a35b-instruct

Reasoning/Thinking

Models with enhanced reasoning:

• kimi-k2-thinking

Troubleshooting

API Key Issues

Problem: "Invalid API key" error

Solution:

1. Verify your API key in AWS Bedrock Console
2. Run: Manage AWS Bedrock → "Clear API Key (Mantle)"
3. Re-enter your API key

Model Not Available

Problem: "Model not available in region" error

Solution:

• Not all models are available in all regions
• Try changing to us-east-1 (widest availability)
• Check AWS Bedrock Model Availability

Rate Limiting

Problem: "Rate limit exceeded" error

Solution:

• Wait a few moments and try again
• Consider using smaller models for testing
• Check your AWS Bedrock quotas in AWS Console

Connection Issues

Problem: Network or timeout errors

Solution:

• Check your internet connection
• Verify firewall/proxy settings allow access to *.api.aws
• Ensure the selected region is accessible from your location

Development

Building from Source

# Install dependencies
npm install

# Compile TypeScript
npm run compile

# Watch mode for development
npm run watch

# Run linting
npm run lint

Or use the Makefile shortcuts:

make install
make compile
make watch
make lint

Debugging

1. Open the project in VS Code
2. Press F5 to launch Extension Development Host
3. Set breakpoints in source files
4. Test the extension in the new window

Project Structure

bedrock-vscode-chat/
├── src/
│   ├── extension.ts           # Extension entry point
│   ├── provider.ts             # Main provider implementation
│   ├── bedrockNative.ts        # Native Bedrock Converse API
│   ├── externalModelMetadata.ts # External model metadata loader
│   ├── types.ts                # TypeScript type definitions
│   └── utils.ts                # Utility functions
├── package.json                # Extension manifest
├── tsconfig.json               # TypeScript configuration
├── icon.svg                    # Source icon (editable)
├── icon.png                    # Extension icon (128x128)
├── README.md                   # This file
├── CONTRIBUTING.md             # Development guide
└── PLAN.md                     # Architecture details

Contributing

Contributions are welcome! See CONTRIBUTING.md for detailed development guidelines.

Quick start for contributors:

1. Fork and clone the repository
2. Install dependencies: npm install
3. Compile: npm run compile
4. Press F5 to launch Extension Development Host
5. See CONTRIBUTING.md for testing, logging, and publishing guidelines

Key development notes:

• Publisher name: easytocloud (lowercase)
• Use Output Channel for logging, not console.log
• Include node_modules in VSIX (required for AWS SDK)
• Test in both F5 mode and installed VSIX
• Use rsvg-convert for icon generation

Resources

• AWS Bedrock Mantle Documentation
• VS Code Language Model API
• OpenAI API Reference
• Contributing Guide - Detailed development documentation

License

MIT License - See LICENSE file for details

Credits

• Project Lead: easytocloud
• Development Assistant: GitHub Copilot

Acknowledgments

Inspired by the HuggingFace VSCode Chat extension.

Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• AWS Bedrock: AWS Support

---

Version: 0.3.1
Status: Production Last Updated: February 5, 2026