Quickstart

Send your first LLM request through the CoreValue AI Gateway in under 2 minutes.

The CoreValue AI Gateway is an OpenAI-compatible API at https://gateway.corevalue.dev. Point your existing OpenAI SDK at it, use provider-prefixed model names to route to any supported provider, and every request is automatically logged with real cost tracking.

Supported Providers

Use provider/model-name format to route to any provider:

Prefix	Provider	Example
(none)	OpenAI	`gpt-4o-mini`
`anthropic/`	Anthropic	`anthropic/claude-3.5-sonnet`
`google/`	Google Vertex AI	`google/gemini-1.5-pro`
`bedrock/`	AWS Bedrock	`bedrock/anthropic.claude-3.5-sonnet`
`groq/`	Groq	`groq/llama-3.3-70b-versatile`
`together/`	Together AI	`together/meta-llama/Llama-3.3-70B-Instruct-Turbo`
`mistral/`	Mistral AI	`mistral/mistral-large-latest`
`deepseek/`	DeepSeek	`deepseek/deepseek-chat`
`xai/`	xAI (Grok)	`xai/grok-2`
`fireworks/`	Fireworks AI	`fireworks/accounts/fireworks/models/llama-v3p1-70b-instruct`
`perplexity/`	Perplexity	`perplexity/sonar`
`cerebras/`	Cerebras	`cerebras/llama-3.3-70b`
`deepinfra/`	DeepInfra	`deepinfra/meta-llama/Llama-3.3-70B-Instruct`
`novita/`	Novita AI	`novita/meta-llama/llama-3.3-70b-instruct`
`nebius/`	Nebius	`nebius/meta-llama/Meta-Llama-3.1-70B-Instruct`
`baseten/`	Baseten	`baseten/qwen2.5-72b`
`chutes/`	Chutes AI	`chutes/Llama-3.3-70B`
`openrouter/`	OpenRouter	`openrouter/anthropic/claude-3.5-sonnet`
`canopywave/`	Canopywave	`canopywave/llama-3.3-70b`

OpenAI models don't need a prefix — just use the model name directly (e.g. gpt-4o-mini). All other providers require the provider/ prefix.

Sign up for a CoreValue account
Go to API Keys and generate a key — it starts with sk-cova-
Add credits at Credits (0% markup, pay only what providers charge)

Use the OpenAI SDK pointed at the gateway. OpenAI models work with no prefix:

import { OpenAI } from "openai";

const client = new OpenAI({
  baseURL: "https://gateway.corevalue.dev/v1",
  apiKey: process.env.COVA_API_KEY, // sk-cova-XXXXXXXXXXXXXXXX
});

const response = await client.chat.completions.create({
  model: "gpt-4o-mini",
  messages: [{ role: "user", content: "Hello, world!" }],
});

console.log(response.choices[0].message.content);

from openai import OpenAI

client = OpenAI(
    base_url="https://gateway.corevalue.dev/v1",
    api_key=os.environ["COVA_API_KEY"],  # sk-cova-XXXXXXXXXXXXXXXX
)

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Hello, world!"}],
)

print(response.choices[0].message.content)

curl https://gateway.corevalue.dev/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $COVA_API_KEY" \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [{"role": "user", "content": "Hello, world!"}]
  }'

Your request appears in the Requests dashboard within seconds, with real cost, latency, and provider info tracked automatically.

Route to any provider by adding the provider/ prefix:

// Anthropic
const response = await client.chat.completions.create({
  model: "anthropic/claude-3.5-sonnet",
  messages: [{ role: "user", content: "Hello!" }],
});

// Groq (fast inference)
const response2 = await client.chat.completions.create({
  model: "groq/llama-3.3-70b-versatile",
  messages: [{ role: "user", content: "Hello!" }],
});

# Anthropic
response = client.chat.completions.create(
    model="anthropic/claude-3.5-sonnet",
    messages=[{"role": "user", "content": "Hello!"}],
)

# Groq (fast inference)
response2 = client.chat.completions.create(
    model="groq/llama-3.3-70b-versatile",
    messages=[{"role": "user", "content": "Hello!"}],
)

# Anthropic
curl https://gateway.corevalue.dev/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $COVA_API_KEY" \
  -d '{"model": "anthropic/claude-3.5-sonnet", "messages": [{"role": "user", "content": "Hello!"}]}'

The gateway translates the request format automatically — you send OpenAI-format requests, and it handles the provider-specific conversion for Anthropic, Google, and Bedrock.

Install the @cova/gateway SDK to attach Cova-* headers for user tracking, custom properties, and fallback chains:

npm install @cova/gateway

import { OpenAI } from "openai";
import { buildHeaders } from "@cova/gateway";

const client = new OpenAI({
  baseURL: "https://gateway.corevalue.dev/v1",
  apiKey: process.env.COVA_API_KEY,
});

const response = await client.chat.completions.create({
  model: "gpt-4o-mini",
  messages: [{ role: "user", content: "Hello!" }],
  extra_headers: buildHeaders({
    userId: "user_abc123",
    properties: { feature: "chat", tier: "pro" },
    fallbacks: ["groq/llama-3.3-70b-versatile"],
  }),
});

pip install cova-gateway

from openai import OpenAI
from cova_gateway import build_headers

client = OpenAI(
    base_url="https://gateway.corevalue.dev/v1",
    api_key=os.environ["COVA_API_KEY"],
)

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Hello!"}],
    extra_headers=build_headers(
        user_id="user_abc123",
        properties={"feature": "chat", "tier": "pro"},
        fallbacks=["groq/llama-3.3-70b-versatile"],
    ),
)

If gpt-4o-mini fails, the gateway automatically retries with groq/llama-3.3-70b-versatile. See the SDK Quick Start for streaming, native fetch, and response metadata parsing.

What You Get

Every request through the gateway includes:

Real cost tracking — Cova-Request-Cost response header shows the actual cost in USD (not an estimate)
Provider and model info — Cova-Provider and Cova-Model response headers
Credits remaining — Cova-Credits-Remaining response header
Request ID — Cova-Id response header for correlation with the dashboard
Gateway mode — Cova-Gateway-Mode shows passthrough or translated
Automatic logging — every request appears in the Requests dashboard with cost, latency, tokens, and status

Check Request Metadata

Retrieve metadata for any request using the meta endpoint:

curl https://gateway.corevalue.dev/gateway/v1/meta/$REQUEST_ID \
  -H "Authorization: Bearer $COVA_API_KEY"

{
  "requestId": "550e8400-e29b-41d4-a716-446655440000",
  "provider": "openai",
  "model": "gpt-4o-mini",
  "status": "success",
  "requestCost": 0.00015,
  "creditsRemaining": 99.85
}

Or use the SDK:

import { parseGatewayMeta } from "@cova/gateway";

const cova = parseGatewayMeta(response.headers);
console.log(cova.requestCost);   // 0.00015
console.log(cova.provider);      // "openai"
console.log(cova.requestId);     // "550e8400-..."

List Available Models

curl https://gateway.corevalue.dev/v1/models \
  -H "Authorization: Bearer $COVA_API_KEY"

Returns an OpenAI-compatible model list with all models available through the gateway.

Quickstart

Supported Providers

What You Get

Check Request Metadata

List Available Models

What's Next?

Platform Overview

SDK Quick Start

Provider Routing

Fallback Chains

On this page