主要 AI モデルの API を、
1 つのキー・円建てで。

curl https://api.tokentaro.jp/v1/chat/completions \
  -H "Authorization: Bearer $TOKENTARO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-sonnet-5",
       "messages":[{"role":"user","content":"納品書の文面を丁寧語で作って"}]}'

# pip install openai
from openai import OpenAI

client = OpenAI(
    base_url="https://api.tokentaro.jp/v1",
    api_key=os.environ["TOKENTARO_API_KEY"],
)
res = client.chat.completions.create(
    model="claude-sonnet-5",
    messages=[{"role": "user", "content": "納品書の文面を丁寧語で作って"}],
)
print(res.choices[0].message.content, res.usage.total_tokens)

// npm i openai
import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.tokentaro.jp/v1",
  apiKey: process.env.TOKENTARO_API_KEY,
});
const res = await client.chat.completions.create({
  model: "claude-sonnet-5",
  messages: [{ role: "user", content: "納品書の文面を丁寧語で作って" }],
});
console.log(res.choices[0].message.content, res.usage.total_tokens);

# ~/.zshrc（または ~/.claude/settings.json の env）
export ANTHROPIC_BASE_URL="https://api.tokentaro.jp"
export ANTHROPIC_AUTH_TOKEN="tt_live_7f3a…c91e"   # TokenTaro のキー
export ANTHROPIC_MODEL="claude-sonnet-5"            # 任意

# 起動して /status で接続先を確認
claude

# ~/.codex/config.toml
model = "gpt-5.6"
model_provider = "tokentaro"

[model_providers.tokentaro]
name = "TokenTaro"
base_url = "https://api.tokentaro.jp/v1"
env_key = "TOKENTARO_API_KEY"
wire_api = "responses"

• 設定 → モデルプロバイダーOpenAI-API-compatible
• モデル名claude-sonnet-5
• API Keytt_live_••••••••
• API endpoint URLhttps://api.tokentaro.jp/v1
• Completion modeChat