Documentation

From API key to first request in minutes.

AssetMeld uses an OpenAI-compatible interface, so most existing integrations only need three changes: your API key, the AssetMeld base URL and the model ID.

Base URL

https://api.assetmeld.com/v1

Authentication

Authorization: Bearer am_live_...

Create an API key from the dashboard. The full key is shown only once, so store it somewhere secure.

Environment variable

export ASSETMELD_API_KEY="am_live_..."

Python Quick Start

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["ASSETMELD_API_KEY"],
    base_url="https://api.assetmeld.com/v1",
)
response = client.chat.completions.create(
    model="MODEL_SLUG",
    messages=[
        {"role": "user", "content": "Hello"}
    ],
)
print(response.choices[0].message.content)

Copy a model ID from the Models page and replace MODEL_SLUG.

Switch models without changing SDKs

The same client can call any enabled AssetMeld model. Change the model value and keep the rest of your integration.

model="GPT_MODEL_SLUG"

model="GLM_MODEL_SLUG"

model="KIMI_MODEL_SLUG"

Streaming

stream = client.chat.completions.create(
    model="MODEL_SLUG",
    messages=[
        {"role": "user", "content": "Explain DNS simply"}
    ],
    stream=True,
)

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

Streaming responses are forwarded incrementally using Server-Sent Events, so your application can process output as it arrives instead of waiting for the entire response.

Errors & Request IDs

Every AssetMeld request includes an X-Request-ID. Keep it when contacting support about a failed or incorrectly billed request.

{
  "error": {
    "message": "Insufficient balance.",
    "type": "insufficient_quota",
    "param": null,
    "code": "insufficient_balance"
  }
}

Common error codes:

• invalid_json
• invalid_api_key
• model_not_found
• model_disabled
• insufficient_balance
• rate_limit_exceeded
• proxy_timeout

Prepaid usage billing

AssetMeld uses prepaid USD credit. Each inference request reserves enough balance for the request, executes the model call and then settles the final charge using reported usage. Unused reserved balance is released.

The applicable model price is captured when the request starts.

Adding balance

Top-ups are currently confirmed manually. Follow the PayPal instructions in your dashboard, keep the PayPal transaction ID, then contact AssetMeld on Telegram with your account email and transaction ID.

Balance updates are not automatic during private beta.

Rate limits

Requests are subject to per-key and account limits. When a rate limit is reached, AssetMeld returns HTTP 429 with an appropriate error code.

Retry deliberately after reducing concurrency or waiting for capacity.

Recommended use cases

Coding

Use AssetMeld as the model backend for OpenAI-compatible coding workflows and developer tools.

SaaS

Call the same API from your server-side application to power AI product features.

Agents

Use model IDs as configuration so agents and workflows can switch models without provider-specific integrations.