> ## Documentation Index
> Fetch the complete documentation index at: https://docs.kymaapi.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Aider

> Use Kyma with Aider — AI pair programming in your terminal.

Aider is an AI pair programming tool that works in your terminal. It supports OpenAI-compatible APIs natively.

## Configuration

```bash theme={null}
export OPENAI_API_KEY="ky-your-api-key"
export OPENAI_API_BASE="https://kymaapi.com/v1"
```

Then run:

```bash theme={null}
aider --model openai/qwen-3.6-plus
```

Or for a specific model:

```bash theme={null}
aider --model openai/qwen-3-coder
```

## Recommended Models

| Model               | Alias      | Best For                                                              |
| ------------------- | ---------- | --------------------------------------------------------------------- |
| `qwen-3.6-plus`     | `best`     | **Recommended** — highest quality                                     |
| `deepseek-v3`       | —          | Near-frontier, complex reasoning                                      |
| `qwen-3-32b`        | —          | Fast edits, math                                                      |
| `llama-3.3-70b`     | `balanced` | Great all-rounder                                                     |
| `deepseek-v4-pro`   | —          | 1M context, top reasoning. Complex refactors and long-codebase tasks. |
| `deepseek-v4-flash` | —          | 1M context, value tier. Everyday completions where speed matters.     |

```bash theme={null}
curl "https://kymaapi.com/v1/models/recommend?agent=aider"
```

See all [model aliases](/guides/model-aliases) for shortcuts.

## Notes

* Aider uses the `openai/` prefix for model names: `openai/best` or `openai/qwen-3.6-plus`.
* [Model aliases](/guides/model-aliases) work with Aider: `--model openai/code`.
* For large codebases, use `gemini-2.5-flash` (1M context) with alias `long-context`.
* Kyma's fallback system ensures requests never fail, even during provider outages.

## Prompt Caching

Prompt caching is automatic. Your system prompt and tool definitions are cached for 5 minutes, reducing costs by up to 90% on subsequent requests. No code changes needed. See [Prompt Caching](/guides/prompt-caching) for details.