IDE & Client Setup

What "IDE support" meansCopy Anchor Link

GLM5 exposes an OpenAI-compatible API. Use it inside VS Code or another IDE through a client or coding agent that supports a custom OpenAI-compatible provider.

This documentation does not claim that GLM5 ships a first-party VS Code extension.

Every client needs these three values together:

Verify the key before configuring an IDECopy Anchor Link

curl https://glm5.app/api/v1/models \
  -H "Authorization: Bearer $GLM5_API_KEY"

Then make a small Chat Completions request. If both calls work but the IDE returns 401, check that the client is using a custom provider and that its Base URL is exactly https://glm5.app/api/v1.

OpenAI SDKCopy Anchor Link

import OpenAI from 'openai';

const client = new OpenAI({
  apiKey: process.env.GLM5_API_KEY,
  baseURL: 'https://glm5.app/api/v1',
});

const result = await client.chat.completions.create({
  model: 'glm-5.3-flash',
  messages: [{ role: 'user', content: 'Hello' }],
});

OpenCodeCopy Anchor Link

Use @ai-sdk/openai-compatible, set baseURL to GLM5, and supply your key through an environment variable. Add the GLM5 public model IDs you intend to use.

Oh My Pi (omp)Copy Anchor Link

For omp, define a custom provider in ~/.omp/agent/models.yml:

providers:
  glm5:
    baseUrl: https://glm5.app/api/v1
    api: openai-responses
    apiKey: GLM5_API_KEY
    models:
      - id: glm-5.3-flash
        name: GLM 5.3 Flash
        contextWindow: 1000000
        maxTokens: 8192

Set GLM5_API_KEY in the same shell, then run omp models glm5 to verify the provider. To make it the default model, set modelRoles.default to glm5/glm-5.3-flash in ~/.omp/agent/config.yml.

Pi / Continue / other clientsCopy Anchor Link

Choose the client's custom OpenAI-compatible provider mode. Do not select a built-in OpenAI provider that hardcodes api.openai.com unless the client lets you override the Base URL.

TroubleshootingCopy Anchor Link

After a GLM5 API request reaches the service, responses include X-Request-ID. Open Activity → Logs and search that ID to inspect status, model, key, tokens, credits, latency, and error code without exposing prompt text.