Documentation

Quick start

  1. Open the bot @PalantirGPT_bot and accept the terms.
  2. Top up your balance (the Top-up button).
  3. In the API section create a key. You can view it again there later.
  4. In your client set the Base URL and the key, and put a model name from the price table into the model field.
  5. Send a test message.

Connection details

Base URL
https://api.palantir-gpt.com/v1
API key
sk-… from the bot
Model
glm-5.3-flash
Protocol
OpenAI Chat Completions
Output limit
16384 (set in your client)
Concurrency
up to 3 requests per key

Client setup

Menu names can differ slightly between client versions. The idea is the same everywhere: an “OpenAI Compatible” provider, an address, a key and a model.

Cline

  1. Open the Cline panel and click the settings gear.
  2. API Provider: OpenAI Compatible.
  3. Base URL: https://api.palantir-gpt.com/v1
  4. OpenAI Compatible API Key: your key.
  5. Model ID: glm-5.3-flash
  6. Expand Model Configuration and set Max Output Tokens to 16384.

Roo Code

  1. Open Roo Code settings, Providers.
  2. API Provider: OpenAI Compatible.
  3. Base URL: https://api.palantir-gpt.com/v1
  4. API Key: your key.
  5. Model: glm-5.3-flash
  6. Under Model Configuration set Max Output Tokens to 16384.

Kilo Code

  1. Open Kilo Code settings and choose the OpenAI Compatible provider.
  2. Base URL: https://api.palantir-gpt.com/v1
  3. API Key: your key.
  4. Model: glm-5.3-flash

Cursor

  1. Settings → Models.
  2. Turn on Override OpenAI Base URL and enter https://api.palantir-gpt.com/v1
  3. Paste your key into OpenAI API Key.
  4. Add Custom Model: glm-5.3-flash
  5. Select the model in chat. Features tied to Cursor's own servers may behave differently.

Continue

  1. Open config.yaml and add the model:
models:
  - name: PalantirGPT
    provider: openai
    model: glm-5.3-flash
    apiBase: https://api.palantir-gpt.com/v1
    apiKey: sk-...
    defaultCompletionOptions:
      maxTokens: 16384

Zed

  1. Open settings.json and add the provider:
  2. Enter the key in the provider settings in the Agent Panel.
{
  "language_models": {
    "openai_compatible": {
      "PalantirGPT": {
        "api_url": "https://api.palantir-gpt.com/v1",
        "available_models": [
          { "name": "glm-5.3-flash", "max_tokens": 200000, "max_output_tokens": 16384 }
        ]
      }
    }
  }
}

Aider

  1. Set environment variables and start Aider:
export OPENAI_API_BASE=https://api.palantir-gpt.com/v1
export OPENAI_API_KEY=sk-...
aider --model openai/glm-5.3-flash

OpenCode

  1. Add a provider to opencode.json:
  2. Set the key with /connect or in OpenCode's auth file.
{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "PalantirGPT": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "PalantirGPT",
      "options": { "baseURL": "https://api.palantir-gpt.com/v1" },
      "models": { "glm-5.3-flash": { "name": "glm-5.3-flash" } }
    }
  }
}

ZCode

  1. In the model settings choose a custom OpenAI-compatible provider.
  2. Base URL: https://api.palantir-gpt.com/v1
  3. API Key: your key.
  4. Model: glm-5.3-flash

Other

  1. Find an “OpenAI Compatible”, “Custom OpenAI” or “Custom endpoint” provider in your client.
  2. Base URL: https://api.palantir-gpt.com/v1 (ending in /v1, without /chat/completions).
  3. API Key: your key, Model: glm-5.3-flash.

Code examples

cURL
curl https://api.palantir-gpt.com/v1/chat/completions \
  -H "Authorization: Bearer sk-..." \
  -H "Content-Type: application/json" \
  -d '{"model": "glm-5.3-flash", "max_tokens": 1000,
       "messages": [{"role": "user", "content": "Hello"}]}'
Python
from openai import OpenAI

client = OpenAI(base_url="https://api.palantir-gpt.com/v1", api_key="sk-...")
r = client.chat.completions.create(
    model="glm-5.3-flash",
    max_tokens=2000,
    messages=[{"role": "user", "content": "Hello"}],
)
print(r.choices[0].message.content)
Node.js
import OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://api.palantir-gpt.com/v1", apiKey: "sk-..." });
const r = await client.chat.completions.create({
  model: "glm-5.3-flash",
  max_tokens: 2000,
  messages: [{ role: "user", content: "Hello" }],
});
console.log(r.choices[0].message.content);

Errors

401Invalid or revoked key. Copy the key from the bot again.
402Insufficient balance. Top up in the bot.
400Unknown model or a malformed request. Check the model name in the price table.
413Request too large. Shorten the context.
429More than three concurrent requests on one key. Wait and retry.
503The model provider is temporarily unavailable. Retry in a minute.

Not supported

Only Chat Completions and the model list. Clients that need Anthropic Messages (for example Claude Code directly) or the Responses API will not connect. Embeddings and autocomplete are not available either.