Pi - Replicas Docs

Replicas supports Pi as an SDK-backed coding agent. Pi runs inside the workspace VM, uses Aster or OpenRouter for inference, and persists its session and tool history across turns.

Setup

Pi and Opencode share Aster credentials. DeepSeek Harness, Pi, and Opencode share OpenRouter credentials. Aster is preferred when both are connected.

Create a key in the Aster console or on the OpenRouter keys page.
Paste it under Pi → Aster API Key or Pi → OpenRouter API Key in Organization → Coding Agents. The Aster credential also appears under Opencode; the OpenRouter credential appears under DeepSeek Harness and Opencode.
Set organization credentials for the team, or add a personal key under Personal to override the organization default.

See Credentials for credential scope and fallback behavior.

Models

With Aster connected, Pi provides kimi-k3, glm-5.2, gpt-oss-120b-fast, and gpt-oss-120b. With OpenRouter, Pi offers the models enabled under Pi → OpenRouter models in Coding Agents settings. Organization admins pick these from the connected OpenRouter account, and the selection is shared with DeepSeek Harness and Opencode. Until it is changed, Pi provides:

Model ID	Label	Notes
`z-ai/glm-5.2`	GLM 5.2	Default
`stealth/ox-alpha`	Ox Alpha	Free stealth reasoning model
`deepseek/deepseek-v4-pro-0813`	DeepSeek V4 Pro	Flagship V4 coding model
`deepseek/deepseek-v4-flash-0731`	DeepSeek V4 Flash	Faster V4 model
`minimax/minimax-m3`	MiniMax M3	Long-context agentic coding
`xiaomi/mimo-v2.5-pro`	MiMo V2.5 Pro	Agent-focused Xiaomi model
`moonshotai/kimi-k2.6`	Kimi K2.6	Multimodal coding model

Choose an organization default under Defaults, a personal default in account preferences, or a model for the active chat from the workspace composer.

Using Pi in Workspaces

Choose Pi from the composer’s harness/model picker when creating a chat. Pi can:

Read, write, and edit repository files
Run shell commands with Replicas command protection
Accept image attachments
Persist sessions and tool history across workspace sleep
Discover provisioned and repository skills through Pi’s global .agents/skills directory
Run skills explicitly with /skill:<name>

Pi does not currently provide native plan-mode, fast-mode, goal-mode, or environment MCP support. Ask Pi for a read-only implementation plan in the prompt when you want planning without changes.

Programmatic Use

Select Pi anywhere Replicas accepts a coding agent:

Surface	How to choose Pi
CLI	`-a pi` on `replicas create` and `replicas send`
API	`"coding_agent": "pi"` on create and message requests
MCP	`"coding_agent": "pi"` on `create_replica` and `send_replica_message`
Automations	Set Coding agent to Pi in the automation editor
Defaults	Select Pi as the organization or personal default coding agent

replicas create fix-tests \
  -e backend \
  -a pi \
  -m "Inspect the repository and fix the failing tests"

curl -X POST "https://api.replicas.dev/v1/replica" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "fix-tests",
    "environment_id": "ENVIRONMENT_UUID",
    "message": "Inspect the repository and fix the failing tests",
    "coding_agent": "pi",
    "model": "z-ai/glm-5.2"
  }'

Get started

Workspaces

Coding agents

Integrations

Build and automate

Apps

Admin

Updates

Pi

Setup

Models

Using Pi in Workspaces

Programmatic Use

​Setup

​Models

​Using Pi in Workspaces

​Programmatic Use

Setup

Models

Using Pi in Workspaces

Programmatic Use