Quickstart

Connect in minutes.

Create a key in your dashboard, add funds with a card, and use the compatibility mode your application already speaks — or drop the base URL into any of the tools below.

Model names

These are the model identifiers we serve. Use one as the model parameter, as ANTHROPIC_MODEL, or in your client's model switcher (for example /model claude-sonnet-5 in Claude Code). Per-token prices are on the pricing table.

Model	Identifier
Claude Opus 5	`claude-opus-5`
Claude Sonnet 5	`claude-sonnet-5`
Claude Fable 5	`claude-fable-5`
GPT-5.6 Sol	`gpt-5.6-sol`
GPT-5.6 Terra	`gpt-5.6-terra`
GPT-5.6 Luna	`gpt-5.6-luna`
Gemini 3.6 Flash	`gemini-3.6-flash`
Grok 4.5	`grok-4.5`
Kimi K3	`kimi-k3`

Anthropic / Claude Code

export ANTHROPIC_BASE_URL=https://undercut.pro
export ANTHROPIC_API_KEY=gw_your_key
export ANTHROPIC_MODEL=claude-opus-5

claude

# Switch model inside a session (any identifier from the table above):
# /model claude-sonnet-5

OpenAI SDK

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "gw_your_key",
  baseURL: "https://undercut.pro/v1",
});

const response = await client.chat.completions.create({
  model: "gpt-5.6-sol",
  messages: [{ role: "user", content: "Hello" }],
});

Cursor

// Cursor Settings → Models → OpenAI API Key
Base URL:  https://undercut.pro/v1
API Key:   gw_your_key
Model:     gpt-5.6-sol   // add as a custom model name

Cline (VS Code)

// VS Code → Cline extension → API Provider: OpenAI Compatible
Base URL:  https://undercut.pro/v1
API Key:   gw_your_key
Model:     gpt-5.6-sol

Codex CLI

export OPENAI_BASE_URL=https://undercut.pro/v1
export OPENAI_API_KEY=gw_your_key

codex --model gpt-5.6-sol

LangChain (Python)

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    base_url="https://undercut.pro/v1",
    api_key="gw_your_key",
    model="gpt-5.6-sol",
)

n8n

// n8n → Credentials → OpenAI API
Base URL:  https://undercut.pro/v1
API Key:   gw_your_key
Model:     gpt-5.6-sol
// Use in any "OpenAI" or "OpenAI Chat Model" node.

Batch discount

Tag non-interactive, latency-tolerant requests as batch traffic to get a 50% discount on every token.

curl https://undercut.pro/v1/chat/completions \
  -H "Authorization: Bearer gw_your_key" \
  -H "X-Undercutai-Protocol: batch" \
  -d '{ "model": "gpt-5.6-sol", "messages": [...] }'
// Batch-tagged requests are billed at a 50% discount. See the calculator below.

Cost calculator

Estimate the cost of a request before you send it.

ModelInput tokensOutput tokens

$5.0625

Estimate only — actual cost also depends on cache usage and batch mode.

Billing note

Dashboard costs are derived from the configured per-million input and output token rates. They are recorded with each usage item for deterministic history.