Codex

Codex is OpenAI’s official AI coding agent for the desktop.

Install Codex

You can skip Codex CLI installation when using the one-click setup wizard below. If it cannot find the codex command, it offers to install the official @openai/codex npm package. Download and install the Codex desktop app from the OpenAI Codex page.

Configure MiniMax API

Option 1: One-click setup wizard
Option 2: Manual setup

Run:

npx -y mmx-cli@latest agent setup

Select Codex. If Codex CLI is missing, keep Codex selected in the second multi-select. The wizard installs the official npm package, verifies its version, configures ~/.codex/config.toml, and creates ~/.codex/mmx-model-catalog.json with the MiniMax model capabilities.Restart Codex after setup. See One-click setup wizard for command options and backup behavior.

Edit the config file

Open ~/.codex/config.toml and add the following, replacing <MINIMAX_API_KEY> with your key from the MiniMax Developer Platform:

model = "MiniMax-M3"
model_provider = "minimax"
model_context_window = 1000000

[model_providers.minimax]
name = "MiniMax"
base_url = "https://api.minimax.io/v1"
experimental_bearer_token = "<MINIMAX_API_KEY>"
wire_api = "responses"

Restart Codex and start using MiniMax-M3

Restart Codex — MiniMax-M3 is now ready to use.

Configure Model Capabilities (Optional)

Skip this section if you used the one-click setup wizard above. Codex can use a custom model catalog to recognize MiniMax-M3 capabilities such as multimodal input, reasoning effort as the thinking switch, system prompt instructions, tool type, and other detailed model parameters. After this is configured, type /model in Codex CLI to see MiniMax-M3 and its available reasoning levels in the model list. Add the following line to ~/.codex/config.toml:

model_catalog_json = "~/.codex/model-catalogs/custom-catalog.json"

Then create ~/.codex/model-catalogs/custom-catalog.json with the detailed model configuration:

{
  "models": [
    {
      "slug": "MiniMax-M3",
      "display_name": "MiniMax-M3",
      "description": "MiniMax",
      "default_reasoning_level": "high",
      "supported_reasoning_levels": [
        { "effort": "none", "description": "Think-Off" },
        { "effort": "high", "description": "Deep" }
      ],
      "shell_type": "shell_command",
      "visibility": "list",
      "supported_in_api": true,
      "priority": 0,
      "base_instructions": "You are Codex, a coding agent based on MiniMax-M3. You and the user share the same workspace and collaborate to achieve the user's goals.",
      "supports_reasoning_summaries": true,
      "default_reasoning_summary": "none",
      "support_verbosity": false,
      "truncation_policy": { "mode": "bytes", "limit": 10000 },
      "supports_parallel_tool_calls": true,
      "experimental_supported_tools": [],
      "input_modalities": ["text", "image"]
    }
  ]
}

Common fields in this configuration:

slug / display_name: The model identifier and display name used in Codex config and the /model list. Keep this aligned with the model name used by the API.
default_reasoning_level: The default reasoning effort. For MiniMax-M3, any non-none value enables Adaptive Thinking; the value does not tune reasoning depth.
supported_reasoning_levels: Reasoning options users can switch between in /model. none turns thinking off; high enables Adaptive Thinking.
base_instructions: Base system prompt Codex adds when using this model. Use it to describe the model identity and collaboration style.
supports_reasoning_summaries: Enables Codex’s Responses API reasoning path for this model. Set this to true so Codex sends reasoning.effort; otherwise Codex omits the reasoning field even when default_reasoning_level is configured. In this example, default_reasoning_summary is set to none so Codex does not request a separate reasoning summary.
shell_type: Declares the shell tool-call type supported by the model. This example uses shell_command.
visibility / supported_in_api / priority: Control whether the model appears in the list, whether it is available through the API, and its ordering priority in the model list.
supports_parallel_tool_calls: Indicates that the model supports parallel tool calls, allowing Codex to handle multiple tool requests.
experimental_supported_tools: Reserved list for experimental tool capabilities. Leave it as an empty array when no extra tools are needed.
input_modalities: Supported input modalities. ["text", "image"] means text and image input are supported.
truncation_policy: Controls truncation behavior for retained context or tool output. This example limits retained content by bytes.

Restart Codex after changing the model catalog.

Token Plan

Use M3 in AI Coding Tools

MCP Guide

CLI Guide

Best Practices

Install Codex

Configure MiniMax API

Configure Model Capabilities (Optional)

​Install Codex

​Configure MiniMax API

​Configure Model Capabilities (Optional)

Install Codex

Configure MiniMax API

Configure Model Capabilities (Optional)