OpenAI-Compatible API

from openai import OpenAI

# ── One-line swap: point the client at Tapas ─────────────────────
client = OpenAI(
    api_key="your-tapas-api-key",   # Your Tapas API key
    base_url="https://tapas.one/v1", # Tapas OpenAI-compatible endpoint
)

# ── Standard chat completion — identical to OpenAI usage ─────────
response = client.chat.completions.create(
    model="tapas-lem",   # "tapas-lem" for cache mode, "tapas-llm" for full inference
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user",   "content": "How does quantum computing work?"},
    ],
)

print(response.choices[0].message.content)
# → "• Quantum computers use qubits instead of classical bits
#    • Qubits exploit superposition to represent 0 and 1 simultaneously
#    ..."

# ── Energy metadata is returned in the response extras ───────────
extras = response.model_extra or {}
print(f"Mode:    {extras.get('tapas_mode', 'unknown')}")
print(f"Saved:   {extras.get('energy_wh_saved', 0):.3f} Wh")
print(f"Latency: {extras.get('response_time_ms', 0)} ms")

from openai import OpenAI

client = OpenAI(
    api_key="your-tapas-api-key",
    base_url="https://tapas.one/v1",
)

# ── Streaming response ────────────────────────────────────────────
stream = client.chat.completions.create(
    model="tapas-llm",  # Full inference mode for streaming prose
    messages=[{"role": "user", "content": "Explain CRISPR gene editing."}],
    stream=True,
)

for chunk in stream:
    delta = chunk.choices[0].delta.content or ""
    print(delta, end="", flush=True)
print()  # newline at end

import OpenAI from "openai";

// ── One-line swap: point the client at Tapas ─────────────────────
const client = new OpenAI({
  apiKey:  "your-tapas-api-key",
  baseURL: "https://tapas.one/v1",
});

// ── Standard chat completion ──────────────────────────────────────
const response = await client.chat.completions.create({
  model: "tapas-lem",
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user",   content: "What is compound interest?" },
  ],
});

console.log(response.choices[0].message.content);
// → "• Compound interest is interest calculated on both the initial
//    principal and the accumulated interest from previous periods..."

import OpenAI from "openai";

const client = new OpenAI({
  apiKey:  "your-tapas-api-key",
  baseURL: "https://tapas.one/v1",
});

// ── Streaming response ────────────────────────────────────────────
const stream = await client.chat.completions.create({
  model:    "tapas-llm",
  messages: [{ role: "user", content: "How does TCP/IP work?" }],
  stream:   true,
});

for await (const chunk of stream) {
  const delta = chunk.choices[0]?.delta?.content ?? "";
  process.stdout.write(delta);
}
console.log();

from langchain_openai import ChatOpenAI
from langchain_core.messages import HumanMessage, SystemMessage

# ── Drop-in swap: replace ChatOpenAI base_url ─────────────────────
tapas = ChatOpenAI(
    model="tapas-lem",
    openai_api_key="your-tapas-api-key",
    openai_api_base="https://tapas.one/v1",
    temperature=0,
)

# ── Use exactly as you would ChatOpenAI ──────────────────────────
messages = [
    SystemMessage(content="You are a helpful assistant."),
    HumanMessage(content="What causes inflation?"),
]

response = tapas.invoke(messages)
print(response.content)

# ── Works with chains, agents, and RAG pipelines ─────────────────
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.output_parsers import StrOutputParser

chain = (
    ChatPromptTemplate.from_template("{question}")
    | tapas
    | StrOutputParser()
)

result = chain.invoke({"question": "How does mRNA vaccine technology work?"})
print(result)

import { ChatOpenAI } from "@langchain/openai";
import { HumanMessage, SystemMessage } from "@langchain/core/messages";

// ── Drop-in swap: replace configuration ──────────────────────────
const tapas = new ChatOpenAI({
  modelName:    "tapas-lem",
  openAIApiKey: "your-tapas-api-key",
  configuration: {
    baseURL: "https://tapas.one/v1",
  },
  temperature: 0,
});

// ── Use exactly as you would ChatOpenAI ──────────────────────────
const response = await tapas.invoke([
  new SystemMessage("You are a helpful assistant."),
  new HumanMessage("What is the Higgs boson?"),
]);

console.log(response.content);