bn-router

Mercury 3 API

Call Inception Mercury 3 through bnrouter’s OpenAI-compatible API. Facts from Inception’s 19 October 2026 product blog and official API docs; current price and model id are in the console.

Inception · model page

What Mercury 3 is

Inception’s product blog dated 19 October 2026 introduces Mercury 3 as a diffusion-based LLM. Inception says it is available on the Inception API Platform and through OpenRouter, Baseten, and Vercel, and that the official API is OpenAI-compatible chat completions.

This page covers the Mercury 3 chat model only. It does not claim that bnrouter already proxies other Inception products, and it does not invent throughput numbers or a gateway model id or price.

Official announcement: Read Inception’s Mercury 3 post (19 October 2026)

Where developers can use it

Inception lists these official surfaces first. bnrouter is an optional OpenAI-compatible gateway in front of connected upstream accounts.

  1. Inception API Platform (vendor id mercury-3), as documented in Inception’s official API docs
  2. OpenRouter listing inception/mercury-3
  3. Partners named in the 19 October 2026 post: OpenRouter, Baseten, and Vercel
  4. bnrouter, if the operator has connected an upstream that serves this model. The plaza may still show coming soon — confirm the id in the console or GET /v1/models.

How to call it through bnrouter

bnrouter accepts the OpenAI chat completions protocol. Create an API key in the console, then send requests to the gateway base URL. This repository does not yet record a gateway model id — copy the id shown in the console (or from GET /v1/models) into the model field. The plaza listing may still show coming soon.

curl
curl https://bnrouter.com/v1/chat/completions \
  -H "Authorization: Bearer $YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "YOUR_MODEL_ID",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
Python (OpenAI SDK)
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://bnrouter.com/v1",
)

resp = client.chat.completions.create(
    model="YOUR_MODEL_ID",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(resp.choices[0].message.content)

Replace YOUR_API_KEY and YOUR_MODEL_ID with values from the console. Do not assume a model id that is not listed there.

Pricing

See current pay-as-you-go price after you register at https://console.bnrouter.com. This page does not list a gateway price.

Questions

What is Mercury 3?

Inception’s 19 October 2026 product blog describes Mercury 3 as a diffusion-based LLM with an official OpenAI-compatible chat completions API.

Is Mercury 3 available on bnrouter?

Inception said Mercury 3 is live on the Inception API Platform and listed OpenRouter, Baseten, and Vercel as distribution surfaces. On bnrouter, confirm the plaza or GET /v1/models — the listing may still show coming soon. See also the models index, quickstart, pricing, and API reference on this site.

How do I call Mercury 3 through bnrouter?

Create a key in the console and POST to the OpenAI-compatible /v1/chat/completions endpoint on the gateway. Use the model id shown in the console — this repo does not yet list one. The plaza may still show coming soon.

What model id should I send?

Use the identifier listed for this model in the console or returned by GET /v1/models. A gateway model id is not configured in this repository. Do not treat the vendor id mercury-3 as a bnrouter id.

How much does the Mercury 3 API cost on bnrouter?

See current pay-as-you-go price after you register at https://console.bnrouter.com. This page does not list a gateway price.

Related pages

Try it in the console

Create an API key, confirm the model id in the model plaza, and send a first request.

Open console