Short answer
Meta released Muse Ember (muse-ember) on November 9, 2026, describing it as its smallest and most cost-efficient model yet. The launch materials give AI Gateway as the way in, with the model written meta/muse-ember — meta/ is the vendor prefix the gateway adds; the model ID itself is muse-ember.
This is not the kind of model where step one is signing up with the vendor. This cycle announces no Meta developer sign-up, no price list and no available-regions list. The official route's gates sit on the gateway's account and payment, not on a Meta account. So:
- "No official account" is not an extra obstacle here — there is no such account to apply for in the materials.
- What you do need is a gateway account and a payment method it accepts.
- With no published API rate, you cannot budget in advance; the materials also mention no free tier.
muse-ember is not on TeamoRouter yet. Once it lands you will not need a Meta account or an overseas card — the same key with model set to muse-ember is enough. There is no Meta model on TeamoRouter today, so no same-vendor stand-in is recommended.
Where the gates are
Both on the gateway side:
- Account: the documented path is AI Gateway, so calling the model means having an AI Gateway account.
- Payment: AI Gateway passes provider pricing through with no markup and charges no platform fee on inference. That says the gateway adds nothing on top — it does not make the model cheap, because Meta has not published a rate at all. Payment follows the gateway's rules, so if you do not hold an overseas card, confirm that step before you plan around it.
Coding agents take the same road: install the latest Vercel CLI, run npx vercel ai-gateway setup, then pick meta/muse-ember in the agent. That still lands on a gateway account.
China and Russia: what the materials say, and what they do not
Straight answer: this launch cycle includes no available-regions list for Muse Ember and no country-specific payment note. So this article does not declare "works in China" or "does not work in Russia" — there is no basis for either in the sources.
What is confirmed is the shape of the route: the official entry point is AI Gateway, not a Meta console. Whether you can use it therefore depends on opening a gateway account and paying by its rules, not on holding an official account in some region.
Whether the gateway endpoint is reachable from your network is a separate question that the launch materials do not address. The fastest way to settle it is to send a real request and see.
Two routes compared
| Route | Meta account | AI Gateway account | Payment the gateway accepts | Usable today |
|---|---|---|---|---|
| Documented path, AI Gateway | No | Yes | Yes | Yes, at an unpublished rate |
| TeamoRouter (once listed) | No | No | No — top up in the console | Not yet, model not listed |
No overseas credit card?
- Going the official route now: you are limited to whatever the gateway accepts; the materials offer no alternative.
- Waiting for TeamoRouter: top up the console from $5, cards supported, billed by the tokens you actually use. No overseas card. Details under "How to pay" below.
What you can do right now
muse-ember has not entered the catalog, but two things are worth doing ahead of it:
- Put the model name behind a config value, so launch day is a one-line change.
- Open the TeamoRouter account and top up, from $5 — then you can run it the day it lands.
This article names no stand-in, for the same reason the API article doesn't: TeamoRouter carries no Meta model, and a lightweight model from another vendor differs in context length, tool-calling conventions and reasoning-effort defaults. An agent tuned for Muse Ember will likely need re-tuning — that is not a model-name swap.
How to call it once it lands
from openai import OpenAI
client = OpenAI(base_url="https://api.teamorouter.com/v1", api_key="YOUR_KEY")
resp = client.chat.completions.create(
model="muse-ember",
messages=[{"role": "user", "content": "Explain what a context window is in one sentence."}],
)
print(resp.choices[0].message.content)
muse-ember is not on TeamoRouter yet. Once it is live, the same key and model: muse-ember is all you change; there is no Meta model on the catalog today, so there is no substitute to recommend.
How to pay
No subscription: top up your balance in the console from $5 and each request is billed by the tokens it uses. Cards are supported, and so are USDT, USDC and ETH. The same key also calls Claude and GPT models.
New accounts start with a $0 balance — top up before your first request. Top up from $5. If a request returns 402 (400 on the OpenAI format) with "wallet balance is insufficient", your balance is empty — your base URL and key are fine. Top up and retry.
FAQ
Can I use Muse Ember directly from China?
This launch cycle has no available-regions list and no China-specific note, so I can't give a yes or no on Meta's behalf. What is confirmed: the official route needs an AI Gateway account and a payment method it accepts, not a Meta account. Once TeamoRouter lists it, all you need is a TeamoRouter key.
Does it work from Russia?
Same picture — no region list and no payment note in the sources. The official route turns on a gateway account and paying by the gateway's rules, and the materials name no country restrictions. After it lands, a TeamoRouter key covers it without any of that.
What if I have no official account?
There is no "official account" sign-up for Muse Ember in this cycle — the documented route wants an AI Gateway account. So what you lack is a gateway account, not a Meta one. If you would rather not deal with either, wait for TeamoRouter to list it and use our key.
What if I have no overseas credit card?
On the official route today you follow the gateway's payment rules, and the materials give no other option. Once Muse Ember is on TeamoRouter, you top up the console from $5 with a card — no overseas credit card needed.
Has Meta published a price?
No. There is no API rate in this cycle, only the relative claim that it delivers more intelligence per dollar than Claude Haiku 5.5 or GPT-6 Luna. See Muse Ember API for where pricing stands.
Is there a free tier?
The materials mention no free tier and no API rate. Do not plan around a free allowance.
Can I use it on TeamoRouter now?
No — muse-ember has not entered our catalog. Once it does, same key, same base URL, only model changes.
Will I need to change the endpoint after launch?
No. The documented usage is just the model name meta/muse-ember, with no separate endpoint behind it. On TeamoRouter it works the same way: change model.
Related reading
- Muse Ember API is out: the model ID, pricing status and how to call it
- Muse Ember vs Claude Haiku 5.5 vs GPT-6 Luna: the full comparison
- Vercel AI Gateway announcement: the official source