Skip to content
OfferSign up and get $10 of credit for any model in Chat and the API. Top up from $1, with no processing fees. See free models →·Some free models are at the daily cap — refills 05:30 IST; requests auto-route to a live backup. Status →
One AI API Key

Meta · Muse · Flagship

Muse Wonder

UnavailableDiscount −30%New

Meta's frontier reasoning model for complex agent tasks, with text, image, audio and video input.

Requests to this model fail right now: it has no working upstream. The rates below are what it will cost when it returns.

Model id

muse-wonder

Price per 1M tokens

Input
Regular rate $4.00now $2.80
Output
Regular rate $20.00now $14.00
Cached input
$0.2093% below input

30% below the market rate.

Added here 18 Oct 2026 — read the changelog for what changed and why.

Output is 5× the input rate. $1 covers about 102 prompts of 1,000 input + 500 output tokens. 1 credit = $1 of usage. All rates

Start using Muse Wonder

Set ONEAIAPIKEY_API_KEY to a key from your console, then run any example. Streaming is on so long answers are not cut off by idle-connection timeouts.

curl https://api.oneaiapikey.com/v1/chat/completions \
  -H "Authorization: Bearer $ONEAIAPIKEY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "muse-wonder",
    "stream": true,
    "messages": [{"role": "user", "content": "List every obligation the supplier has in this contract, with the clause number for each."}]
  }'
import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.oneaiapikey.com/v1",
    api_key=os.environ["ONEAIAPIKEY_API_KEY"],
)

stream = client.chat.completions.create(
    model="muse-wonder",
    messages=[{"role": "user", "content": "List every obligation the supplier has in this contract, with the clause number for each."}],
    stream=True,  # recommended for long outputs
)
for chunk in stream:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)
import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.oneaiapikey.com/v1",
  apiKey: process.env.ONEAIAPIKEY_API_KEY,
});

const stream = await client.chat.completions.create({
  model: "muse-wonder",
  messages: [{ role: "user", content: "List every obligation the supplier has in this contract, with the clause number for each." }],
  stream: true,
});

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

More: quickstart · API reference · try it in Chat first

Benchmarks reported by the vendor

Scores for Muse Wonder as its maker published them, with the setting they state. We have not re-run them. Versions and settings differ between vendors, so compare only identical benchmark names; compare models side by side.

Tokens per day, by model

Last 7 days, UTC days

Hover, tap or drag across any day, or focus the chart and use the arrow keys, to see how every model did that day.

Real tokens served through this gateway, updated every 5 minutes. Ranked by usage, not quality.

Full ranking insights

Muse line-up

Muse Wonder or Muse Spark Open or Claude Opus 5.5 (free lane)?

Of the 2 Muse models on this site, Muse Wonder is the top of the line-up and Muse Spark Open the step down. Claude Opus 5.5 (free lane) from Anthropic is the closest-priced flagship-tier model from another vendor. Each comparison lists the published rates and limits, then what favours each side.

Muse WondervsMuse Spark Open

Per 1M tokensMuse WonderMuse Spark Open
Input$2.803.2× as much$0.875
Output$144.7× as much$2.98
Cached input$0.2033% more$0.15
Context1.05M0% larger1.05M

Pick Muse Wonder when

  • the task needs a higher tier: this catalogue places Muse Wonder in the flagship tier and Muse Spark Open in the balanced tier.
  • prompts run past 1.05M tokens: Muse Wonder reads up to 1.05M.
  • the work is agent tasks, multimodal input and long documents, which this catalogue lists for Muse Wonder and not for Muse Spark Open.

Pick Muse Spark Open instead when

  • the job runs well enough on the lower rates: input at 69% less and output at 79% less.
  • your workload looks like coding, agent workflows and tool calling, labels the catalogue gives Muse Spark Open only.

Muse WondervsClaude Opus 5.5 (free lane)

Per 1M tokensMuse WonderClaude Opus 5.5 (free lane)
Input$2.80the same$2.80
Output$14the same$14
Context1.05M5% larger1M

Pick Muse Wonder when

  • prompts run past 1M tokens: Muse Wonder reads up to 1.05M.
  • the work is agent tasks, multimodal input and long documents, which this catalogue lists for Muse Wonder and not for Claude Opus 5.5 (free lane).

Pick Claude Opus 5.5 (free lane) instead when

  • your workload looks like agentic coding, large refactors and complex reasoning, labels the catalogue gives Claude Opus 5.5 (free lane) only.

Use cases

What Muse Wonder is listed for, and what it costs

The catalogue lists Muse Wonder for agent tasks, multimodal input and long documents. Rows with a price show one typical request at the published rates, with the arithmetic shown. The market-rate figure shows the same request at the lowest public price elsewhere. Token counts are examples; your prompts will differ.

Model your own traffic

  1. Agent tasks

    One step of a tool-using agent: 20,000 tokens of instructions, tool results and history, 15,000 of them cached, and 1,000 tokens of tool calls and notes out.

    “Check the order status for each customer in this list and draft a reply to the late ones.”

    Same request on Muse Spark Open $0.0096 and Claude Opus 5.5 (free lane) $0.07.

    $0.031

    per agent step, about 322 per $10

    $0.043 at the market rate

    0.005 × $2.80 + 0.015 × $0.20 + 0.001 × $14

  2. Multimodal input

    The cost depends on the size and number of files you send, so no typical request is priced here.

  3. Long documents

    A long report or contract in one prompt: 100,000 tokens in, a 1,500-token answer out.

    Send whole documents in one request and ask for answers that quote the source.

    “List every obligation the supplier has in this contract, with the clause number for each.”

    Same request on Muse Spark Open $0.091962 and Claude Opus 5.5 (free lane) $0.301.

    $0.301

    per document, about 33 per $10

    $0.43 at the market rate

    0.1 × $2.80 + 0.0015 × $14

Overview

When Muse Wonder fits

Muse Wonder is a flagship model from Meta, available here under the exact model id “muse-wonder”. It heads the Muse line-up here, above Muse Spark Open. Meta has not published a release date we can cite; it has been available here since 18 Oct 2026.

Meta documents tools, vision, reasoning and structured output for it, and it reads images as well as text.

It costs $2.80 per million input tokens and $14 per million output tokens, so output is 5× the input rate. That is 30% below the market rate of $4 input and $20 output per million tokens, the lowest public price we found at other API gateways. On the use cases below, output makes up 45% of the bill for agent tasks and 7% for long documents, so trimming the context you send saves more than capping the answer.

Cached input bills at $0.20 per million tokens, 93% below the standard input rate. In the agent tasks example, 15k of the 20k input tokens come from cache, which brings one agent step from $0.07 down to $0.031.

It is served through the OpenAI-compatible format at https://api.oneaiapikey.com/v1, so any SDK or tool that lets you set a custom base URL, for example Codex CLI, Cursor and Cline, can use it. It is not offered on the Anthropic Messages endpoint; pick an Anthropic-format model for Claude Code.

Use Muse Wonder with any balance.

Muse Wonder is a paid model: any balance above $0 pays for it, the $10.00 sign-up credit included. When it runs out, add $1 or more to your wallet. Every request is billed at the rates above, with a receipt in your usage log.

Muse Wonder is developed by Meta. Model names and trademarks belong to their owners. One AI API Key is an independent API gateway and is not affiliated with, endorsed by or sponsored by Meta.