High-Performance Aider Configuration

A production-optimized Aider configuration that leverages OpenRouter's advanced models with reasoning capabilities, architect mode, and fine-tuned performance settings for enterprise development.

What This Skill Does

This configuration transforms Aider into a high-performance AI coding assistant by:

Using OpenRouter's o4-mini-high model with enhanced reasoning

Enabling architect mode for better code structure planning

Optimizing chat history and token usage

Setting up model aliases for quick switching

Configuring diff-based editing for precise code changes

Instructions

1. Prerequisites

Ensure you have Aider installed and an OpenRouter API key:

```bash

pip install aider-chat

export OPENROUTER_API_KEY="your-key-here"

```

2. Create Configuration File

Create `.aider.conf.yml` in your project root (or `~/.aider.conf.yml` for global config):

```yaml

model: openrouter/openai/o4-mini-high

editor-model: openai/gpt-4.1

weak-model: openai/gpt-3.5-turbo

reasoning-effort: high

edit-format: diff

architect: true

timeout: 60

verify-ssl: true

max-chat-history-tokens: 32000

auto-commits: false

dark-mode: false

cache-prompts: true

no-stream: true

map-tokens: 8192

alias: [fast:gpt-4o-mini, coder:deepseek/deepseek-coder]

```

3. Understanding Key Settings

**Model Configuration:**

`model`: Primary model with reasoning capabilities (o4-mini-high)

`editor-model`: Separate model for code editing tasks (gpt-4.1)

`weak-model`: Lightweight model for simple queries (gpt-3.5-turbo)

`reasoning-effort: high`: Enables deeper analysis for complex problems

**Performance Optimization:**

`max-chat-history-tokens: 32000`: Large context window for complex projects

`map-tokens: 8192`: Increased token allocation for repository mapping

`cache-prompts: true`: Speeds up repeated operations

`timeout: 60`: Extended timeout for complex reasoning tasks

**Development Workflow:**

`architect: true`: Enables planning mode before code changes

`edit-format: diff`: Precise, reviewable code modifications

`auto-commits: false`: Manual control over commits

`no-stream: true`: Complete responses for better review

**Model Aliases:**

`/fast` → gpt-4o-mini (quick iterations)

`/coder` → deepseek-coder (specialized code generation)

4. Usage Examples

**Start Aider with this config:**

```bash

aider

```

**Switch to fast mode mid-session:**

```

/fast

```

**Use specialized coder model:**

```

/coder

```

**Architect mode workflow:**

1. Describe your feature requirement

2. Aider plans the architecture first

3. Review the plan

4. Proceed with implementation

5. Customization Tips

**For faster responses (development):**

```yaml

reasoning-effort: medium

timeout: 30

no-stream: false

```

**For maximum quality (production):**

```yaml

reasoning-effort: high

architect: true

verify-ssl: true

```

**For budget-conscious usage:**

```yaml

model: openrouter/openai/gpt-4o-mini

editor-model: openai/gpt-4o-mini

```

Configuration Reference

| Setting | Value | Purpose |

|---------|-------|---------|

| `reasoning-effort` | high | Deep analysis for complex problems |

| `architect` | true | Plan before coding |

| `edit-format` | diff | Precise, reviewable changes |

| `max-chat-history-tokens` | 32000 | Large context retention |

| `cache-prompts` | true | Performance optimization |

| `auto-commits` | false | Manual commit control |

Important Notes

OpenRouter API key must be set as environment variable

This config prioritizes quality over speed (adjust `reasoning-effort` if needed)

`no-stream: true` means you wait for complete responses

`auto-commits: false` gives you full git control

Model aliases require the specified models to be available via your API

When to Use This Configuration

Enterprise development requiring high code quality

Complex refactoring tasks

Architecture planning and design

Projects with large codebases (32K token context)

When you need reproducible, reviewable changes

When NOT to Use

Rapid prototyping (use `fast` alias instead)

Simple one-line fixes (overhead not needed)

Network-constrained environments (streaming disabled)

Budget-sensitive projects (o4-mini-high is premium)

High-Performance Aider Configuration

High-Performance Aider Configuration

What This Skill Does

Instructions

1. Prerequisites

2. Create Configuration File

3. Understanding Key Settings

4. Usage Examples

5. Customization Tips

Configuration Reference

Important Notes

When to Use This Configuration

When NOT to Use

Reviews (0)