Acceso y precios de la API de Muse Wonder: qué permiten obtener $4/$20 en el primer modelo frontera de Meta

Meta Superintelligence Labs comenzó a vender Muse Wonder el 12 de octubre de 2026, a $4.00 por millón de tokens de entrada y $20.00 por millón de tokens de salida — y la mitad más capaz del lanzamiento, Muse Wonder Full, no tiene ningún precio por token. Esa es toda la historia del acceso en una sola frase: dos checkpoints publicados, dos puertas distintas, y solo una de ellas se abre con una clave de API.

So this page answers the questions that come before any benchmark comparison. Which identifier you can call, what the rate card contains beyond the two headline numbers, what M​eta says the model does, what an outside evaluator has measured so far, and what the current answer is if you wanted to reach it from a router today. Nothing on this page implies OrcaRouter serves Muse Wonder — it does not, and we say so in the section where it matters.

¿Cuál Muse Wonder puedes llamar de verdad?

M​eta's material describes a series rather than a model. The two members behave differently as products, and the difference is administrative rather than technical.

• muse-wonder-1.0 — the checkpoint served on the Meta Model API and inside Muse Code, M​eta's terminal coding agent. This is the identifier a developer provisions against, and every per-token figure below belongs to it.

• muse-wonder-full-1.0 — a separate checkpoint that M​eta describes as the higher-scoring configuration. It is not an effort flag on the API model and it is not a second API endpoint; M​eta sells it through a Muse Code subscription. M​eta's published plans for that product are $50 a month for Power Usage and $200 a month for Maximum Usage, which means Full's real unit of purchase is a seat-month, not a million tokens.

Esa es una forma inusual para un lanzamiento de frontera, y tiene una consecuencia práctica que vale la pena enunciar sin rodeos: no hay ninguna ruta de API hacia el checkpoint con mejor puntuación. Un equipo que evalúa el modelo de la API y le gusta no puede simplemente girar un dial y obtener Full. O compran la suscripción, o aceptan que el modelo que midieron es el modelo que reciben.

La tarjeta de tarifas, línea por línea

M​eta's per-token pricing applies to the API checkpoint. The rows below are vendor rate-card facts unless marked otherwise, all per million tokens.

• Entrada — $4.00 por millón de tokens

• Entrada en caché — 0,20 $ por millón de tokens, lo que el evaluador independiente Artificial Analysis interpreta como un descuento de caché del 95 %

• Salida — $20.00 por millón de tokens

• Costo combinado con una proporción de aciertos de caché/entrada/salida de 7:2:1 — alrededor de $2.94 por millón de tokens, según la propia página del modelo de Artificial Analysis

• Measured cost of completing one Intelligence Index task — about $4.52, again Artificial Analysis' figure rather than M​eta's

• Ventana de contexto — 1,048,576 tokens declarados y servidos

• Salida máxima en una sola respuesta — 943,718 tokens, que es el techo de respuesta única más alto del tablero que seguimos

• Modalidades de entrada — texto, imagen, voz y vídeo, frente a la salida de texto

• Reasoning control — mandatory. M​eta exposes an effort ladder from minimal to max, defaulting to medium, but not an off switch

Un detalle estructural es fácil de pasar por alto y es importante para cualquiera que calcule precios de recuperación o de trabajo con agentes largos: la tarifa de entrada no da un salto en un umbral de tokens. Un prompt de 400.000 tokens se factura a la misma tarifa por token que uno de 400 tokens. Varios modelos insignia de la competencia reajustan el precio de toda la solicitud una vez que un prompt cruza un límite documentado, así que esta es una diferencia real en la forma de la curva, y no solo en su altura.

What M​eta says it scores, and what that is worth

M​eta shipped a benchmark table covering both checkpoints at their maximum reasoning settings, and it is the only scoreboard that covers Muse Wonder Full at all. It is vendor-reported, it was produced in M​eta's own harness, and none of it has been independently reproduced — read the following as M​eta's claim rather than as a measurement.

The rows M​eta promotes are the sustained-agent ones. On SWE-Marathon, a long-horizon software task set, M​eta places Muse Wonder Full ahead of every model in its comparison set. On DeepSWE 1.1 it makes a similar claim, and on the composite index it publishes the series carries its own name. Independent commentary has repeated those two rows selectively, and at least one widely shared post paired a Muse Wonder Full figure with a lower competitor number without noting that the underlying table is the vendor's, unreproduced, and unreleased as a harness.

The awkward part is internal to the table. Muse Wonder Full is not uniformly better than the cheaper API checkpoint: the two profiles trade rows, which suggests M​eta trained toward two different objectives rather than turning one dial up. If that is right, then "Full is the better model" is not a safe assumption for your own workload, and the only way to know which profile fits is to measure both — which currently means buying the subscription.

Lo que ha medido un evaluador externo

Artificial Analysis is the only evaluator with a published row for Muse Wonder as of October 13, 2026, and its numbers are a different species from M​eta's: same harness for every model on the board, published configuration, and no vendor input.

• Índice de Inteligencia — 52 al máximo esfuerzo, noveno en una tabla que abarca más de 200 modelos

• Costo por tarea de indexación — alrededor de $4.52, que el evaluador señala como por encima de la mediana para la cohorte una vez que se incluye el precio

• Tokens de salida consumidos en esa ejecución — unos 130 millones frente a una mediana del ranking cercana a los 85 millones, que el evaluador describe como un tanto verboso

• Velocidad de servicio — aproximadamente 115 tokens de salida por segundo, con un tiempo hasta el primer token cercano a 20 segundos

• Muse Wonder Full — un índice de 56 con esfuerzo máximo, clasificado cerca de lo más alto de la tabla, sin precio por token porque no se vende por token

Note the last row against M​eta's table: on the composite index M​eta publishes 59.0 for Muse Wonder Full while the independent board reads 56. Three points on a scale like that is not a rounding difference, and the independent figure is the one to plan against, because it was produced the same way as the numbers for every other model you are weighing.

Donde todavía no puedes llamarlo

Muse Wonder is available through M​eta's own API and inside Muse Code, and that is currently the whole list. There is no third-party serving coverage in the catalogue we can verify, and it is not in OrcaRouter's catalogue — a router cannot route a checkpoint nobody else is serving, and we would rather say that than imply otherwise.

What that means in practice is a two-vendor setup if you want to try it: M​eta's key for the new model, and whatever you already use for everything else. That is the situation a router is built for on the other side of the fence. OrcaRouter passes provider list price through with 0% markup across more than 200 models behind one key, which includes M​eta's own Muse Spark 1.2 and Muse Spark 1.1 — so the M​eta models you can call today are already reachable in one place, and a rate change at M​eta lands here the same day it is announced rather than waiting for a middleman to re-price. When a second provider does start serving Muse Wonder, automatic failover across endpoints is what keeps a single unhealthy path from taking your traffic down.

¿Qué cambiaría esta respuesta?

Cuatro cosas son verificables, y cualquiera de ellas haría que valiera la pena releer esta página en lugar de volver a conjeturar.

• Un identificador de modelo que aparezca para Muse Wonder Full en la Meta Model API, lo que acabaría con la barrera de solo suscripción

• Una revisión del índice independiente que cambie el 52, o la llegada de un segundo evaluador con su propia fila

• Un cambio de precio en cualquiera de las líneas de la tarifa, que con un margen del 0 % entra en vigor de nuestro lado el mismo día

• Que Muse Wonder aparezca en un catálogo de terceros con una tarifa publicada, lo que convertiría la cuestión de disponibilidad en una cuestión de enrutamiento en lugar de una cuestión de proveedor

Until then the honest summary is that M​eta has priced its API checkpoint 20% under the current A​nthropic and O​penAI flagships on input and output alike, made its cheaper model the one you can reach programmatically, and left the interesting configuration behind a monthly seat. The $4/$20 is real and it is the cheapest frontier rate M​eta has ever published. It is also the price of the model you can actually call.