Skip to main content

Before you start

Complete organization setup, make sure your account has inference credit, and create a key on the API Keys page. Most coding agents need a tool-capable chat model; Aider’s basic edit workflow does not require native tool calling. Image input and reasoning support depend on the model. These terminal examples use Bash or Zsh on macOS, Linux, or Windows through WSL. Install Oh My Pi using its official installation instructions. Its command is omp; it uses a separate configuration directory from Pi. The install script puts omp in ~/.local/bin but does not add that folder to your PATH. If omp --version is not found, add export PATH="$HOME/.local/bin:$PATH" to your shell profile and open a new terminal. Installation instructions.

Get your connection settings

Open Integrations, select a model, then choose Oh My Pi. The screen uses your organization’s region. The examples below use the public global endpoint; replace it with your regional endpoint and replace YOUR_MODEL_ID with the full model ID shown there. Choose one of the models listed for this integration. Set your Pearl API key in the terminal where you run Oh My Pi:

Configure Oh My Pi

Merge the pearl provider into ~/.omp/agent/models.yml. The generated JSON is also valid YAML. PEARL_API_KEY is an environment-variable name in this configuration. Pearl chat models support all three API formats. Choose Responses, Messages, or Chat Completions and use its matching configuration with a model that supports tool calling.
~/.omp/agent/models.yml (JSON is valid YAML; merge providers)
Start Oh My Pi

Choose another API format

Use exactly one provider configuration. Pearl chat models support all three API formats. Choose a format supported by your tool, and replace the provider entry when switching formats. Tool calling, reasoning, and image support vary by model. OpenAI Responses
~/.omp/agent/models.yml (JSON is valid YAML; merge providers)
Start Oh My Pi
Anthropic Messages
~/.omp/agent/models.yml (JSON is valid YAML; merge providers)
Start Oh My Pi

Start using Oh My Pi

Start omp with the explicit Pearl provider and model. The first launch opens a short setup: press Esc to skip provider sign-in, press Enter to keep the Pearl model as the default, then confirm or skip the display steps. Open a project, ask it to explain a file, and then request an edit. Responses configuration uses full history replay with store: false.

Troubleshooting

For Responses, keep statefulResponses false and remove any PI_OPENAI_STATEFUL or OMP_OPENAI_STATEFUL override that enables stored history. Keep remote compaction and provider-hosted features disabled. Reasoning is on in these examples. For Chat Completions, thinkingFormat openrouter sends Pearl’s reasoning object, and qwenPreserveThinking false drops a template switch Pearl rejects. For Messages, the anthropic-adaptive thinking mode replaces thinking budgets, which Pearl rejects. The efforts list is the full ladder so omp never lowers a level; levels the model doesn’t serve round up to the next one it does. The start command sets the level with --thinking, because omp otherwise starts at max. On models that always reason, requiresEffort true sends the lowest level for off, since turning thinking off returns 400.
  • 401: check the key and allow time for a newly created key to reach the inference gateway.
  • 404: copy the endpoint and exact model ID from Integrations. OpenAI base URLs end in /v1; the Anthropic SDK and Claude Code use the URL without that suffix.
  • 402: check your organization’s credit and billing status.
  • 429 or Anthropic 529: respect retry instructions and reduce concurrency.
  • For truncated output, check the output limit and reasoning budget. Increase limits only within the model’s maximum output length.