Kyma APIKyma API
For Roo Code

The best models for Roo Code
Pay per use

Connect Roo Code to Kyma's active open models. Auto-failback, prompt caching, OpenAI-compatible.

Create account Setup Guide

Multi-Mode Ready

Different models for Architect, Code, Edit, and Debug modes. Kyma lets you switch models per mode without switching providers.

Automatic fallback

For eligible failures, Kyma API tries another compatible route when available. Requests can still fail.

Prompt Caching

Repeated context? Kyma caches it, and the repeated part bills at that model's own cached-input rate: listed beside its input rate on every model page.

Recommended Models for Roo Code

Picked for quality, speed, and tool use: matched to Roo Code's multi-mode workflow.

1
Qwen 3.6 PlusVision input

Supports reasoning, tool use, vision input. Compare current prices and measured uptime on the model page.

qwen-3.6-plus

Reasoning

2
Kimi K2.6Vision input

Supports reasoning, tool use, vision input. Compare current prices and measured uptime on the model page.

kimi-k2.6

Reasoning

3
MiniMax M2.5Tool use

Supports reasoning, tool use. Compare current prices and measured uptime on the model page.

minimax-m2.5

Reasoning

4
Qwen 3-32BTool use

Supports reasoning, tool use. Compare current prices and measured uptime on the model page.

qwen-3-32b

Reasoning

5
DeepSeek V4 FlashTool use

Supports reasoning, tool use. Compare current prices and measured uptime on the model page.

deepseek-v4-flash

Reasoning

Setup guide

1. Create account → 2. Install Roo Code from the VS Code Marketplace → 3. Paste the config below.

🦘Roo Code (VS Code)
1. Open VS Code → Extensions → Search "Roo Code" → Install

2. Click Roo Code icon → Settings

3. Set these values:
   API Provider:  OpenAI Compatible
   Base URL:      https://kymaapi.com/v1
   API Key:       kyma-xxxxx  (get yours at kymaapi.com)
   Model ID:      qwen-3.6-plus

Done! Roo Code now uses Kyma's models.
🎯Multi-Mode Setup
Roo Code supports multiple modes. Recommended models per mode:

Code Mode:       kimi-k2.6          (best tool use)
Edit Mode:       qwen-3-32b         (chat)
Debug Mode:      deepseek-v4-flash  (cheap deep analysis)

Tip: Use model aliases for easy switching:
  "best"     → qwen-3.6-plus
  "code"     → qwen-3-coder
  "agent"    → kimi-k2.6
  "fast"     → gemini-3.5-flash-lite

Connect Roo Code to Kyma API

$0.50 signup credit for eligible free-tier models after email verification. No card required.

Create account

Explore more

Qwen 3.6 PlusKimi K2.6MiniMax M2.5Qwen 3-32BDeepSeek V4 FlashCompare modelsLive model rankingsRoo Code docs guide