aitokencounter.org
Token counter Cost calculator Muse Wonder pricing New
Model pricing · MetaNewPreview

Muse Wonder Pricing & Cost Calculator

Prices verified Oct 14, 2026/Price version 2026-10-14.1/Preview — pricing and availability may change/Sources

Muse Wonder, Meta’s frontier model released Oct 12, 2026, costs $4.00 per 1M input tokens and $20.00 per 1M output tokens on the Meta Model API, the same list price as Claude Opus 5.5 and 40% of GPT-6 Astra. Muse Wonder Full is not on the API; it comes only with the Muse Code Power Usage and Maximum Usage plans.

Calculate your cost Compare more models
On this page
  1. 01API pricing
  2. 02Cost calculator
  3. 03Wonder vs Wonder Full
  4. 04Where to get it
  5. 05Cost vs similar models
  6. 06Token estimator
  7. 07FAQ
  8. —Sources & change log

01 — Pricing

Muse Wonder API pricing per 1M tokens

Meta Model API · Standard mode · USDVerified Oct 14, 2026
Inputuncached
$4.00 / 1M
Context window
1,048,576 tokens
Outputincludes reasoning tokens
$20.00 / 1M
Max output
Not verified yet
Cached inputcache read
$0.20 / 1M
Rate limitper team
3,000 RPM · 4,000,000 TPM
Cache writes
No charge · automatic prefix caching
Model ID
muse-wonder-1.0
Long-context premium
None
Status
Preview
Batch
Not available
Released
Oct 12, 2026

Billed per use

Web search grounding
$2.50 / 1,000 queries Not in calculator

02 — Calculator

Cost calculator

Preset to Muse Wonder. Same calculator as the LLM API cost calculator.

Model

Anthropic

Meta

OpenAI

Only models with verified prices are listed. A model is added after we check the vendor’s price page.

Tokens per call
tokens

Everything you send: system prompt, history, tools and user text.

% of input

Share of input tokens read from cache — not share of requests.

tokens

Includes reasoning tokens — don't add them twice.

= 2,000 uncached + 0 cache-read tokens per call. Cache writes: not billed by this vendor.

Volume

API calls

API calls — not users or conversations.

days

The period your total covers.

EstimatePreview

Muse Wonder

Meta Model API · Standard · muse-wonder-1.0

1,000 API calls/day × 30 days

Uncached input$240.002,000 tok × $4.00 / 1M = $0.008000 per call
Cache read—0 tokens
Cache writesn/aNot billed by Meta (automatic prefix caching).
Output · incl. reasoning$1,200.002,000 tok × $20.00 / 1M = $0.04000 per call
Per call$0.04800
Period total30 days · 30,000 API calls$1,440.00
Monthly budget$1,488.00
  • Uncached input16.7%
  • Cache read0.0%
  • Cache write0.0%
  • Output83.3%
  • Max output not verified
  • Not included: web search grounding ($2.50 / 1,000 queries).
  • Excludes tax. Totals are summed before rounding.
Prices verified 2026-10-14Price version 2026-10-14.1USD
Output includes reasoning tokens; at higher reasoning effort the billed output can be several times the visible answer.

03 — Wonder Full

Muse Wonder vs Muse Wonder Full

ComparedMuse WonderMuse Wonder Full
Meta Model APImuse-wonder-1.0 · $4.00 / $20.00 per 1M · Standard mode onlyNot available
Muse CodeAll four plansPower Usage and Maximum Usage only
Muse appComing soon, first to subscribersMeta announcement, Oct 13Not announced
PositioningFast frontier versionFull version

Setting reasoning effort to max on Muse Wonder doesn’t switch it to Wonder Full — they’re separate models.

04 — Plans

Where to get Muse Wonder: API, Muse Code plans, Muse app

ChannelPriceModelsPrice source
Meta Model API
pay as you go		$4.00 in · $20.00 out · $0.20 cached / 1MWonder, Spark 1.1–1.3Meta pricing pageOfficial · checked Oct 14, 2026
Muse Code · Everyday UsageSee Meta's plan pageWonder, SparkShown at checkoutMeta: benefits vary by region
Muse Code · High UsageSee Meta's plan pageWonder, SparkShown at checkoutMeta: benefits vary by region
Muse Code · Power Usage$50 / monthWonder, Wonder Full, Spark Meta models pagePrice: CNBC, Oct 12 (press report)
Muse Code · Maximum Usage$200 / monthWonder, Wonder Full, Spark Meta models pagePrice: CNBC, Oct 12 (press report)
Muse appFree tier · $20 and $100 subscriptionsWonder coming to subscriber tiersMuse app pageFree tier: Meta · $20 / $100: CNBC, Oct 12 (press report)

At API prices, $50 buys 2.5M output tokens or 12.5M input tokens.

A reference point only: Power Usage includes Wonder Full and a fixed allowance, so it isn’t an equivalent conversion.

05 — Compare

Muse Wonder cost vs similar models

Uses the same usage as the calculator above.

Cache on other models
ModelList price / 1M · in · outPer callPeriod · 30 daysvs Muse Wonder
Muse WonderPreviewSelected · Meta$4.00 · $20.00$0.04800$1,440.00Selected
Claude Opus 5.5Anthropic$4.00 · $20.00$0.04800$1,440.00±$0.00 · 0.0%
Claude Sonnet 5.5Anthropic$2.00 · $10.00$0.02400$720.00−$720.00 · −50.0%
Muse Spark 1.3Meta$1.25 · $4.25$0.01100$330.00−$1,110.00 · −77.1%
GPT-6.1 SolOpenAI$2.00 · $10.00$0.02400$720.00−$720.00 · −50.0%
GPT-6 AstraOpenAI$10.00 · $50.00$0.1200$3,600.00+$2,160.00 · +150.0%
  • Default: cache reads are priced as new input on other models — a cache hit rate doesn’t carry over to a model you haven’t run.
  • Each model’s tokenizer is different, so the same text has a different token count. Equal token counts here show a price scenario only.
  • Tool calls such as web search aren’t included in any row.

Compare more models

06 — Estimator

Token estimator

Meta publishes no tokenizer, so this is an estimate: characters ÷ 4.

≈ 0tokensEstimatedAs input on Muse Wonder: about $0.00 per call

Meta’s POST /v1/responses/input_tokens count includes a hidden system prompt Meta adds. Meta says it’s for checking whether a request fits the context window, not for estimating your bill — so we don’t use it here.

07 — FAQ

Muse Wonder FAQ

Is Muse Wonder free?

Not on the API: the Meta Model API charges $4.00 per 1M input tokens and $20.00 per 1M output tokens. Muse Wonder is included in all four Muse Code plans, and Meta says it will come to the Muse app soon, rolling out first to subscribers.

How much does the Muse Wonder API cost?

$4.00 per 1M input tokens, $20.00 per 1M output tokens and $0.20 per 1M cached input tokens. Cache writes are free because caching is automatic, there is no long-context premium, and Batch isn't available. Web search grounding is billed separately at $2.50 per 1,000 queries.

What is the difference between Muse Wonder and Muse Wonder Full?

Muse Wonder is the fast frontier version and the one on the API. Muse Wonder Full is the full version, available only in the Muse Code Power Usage and Maximum Usage plans. Setting reasoning effort to max on Muse Wonder does not switch it to Wonder Full.

Can I use Muse Wonder Full through the API?

No. As of Oct 14, 2026, Wonder Full is not on the Meta Model API. We'll add its API price here once Meta publishes one.

Which Muse Code plan includes Muse Wonder Full?

Power Usage and Maximum Usage. Everyday Usage and High Usage include Muse Wonder but not Full. Meta shows plan prices at checkout and says benefits vary by region; press reports put Power Usage at $50 and Maximum Usage at $200 a month.

What is Muse Wonder's context window?

1,048,576 tokens. We haven't been able to verify a maximum output yet, so the calculator checks requests against the context window only.

Is Muse Wonder available outside the US?

Meta's models page says availability varies by country under its Geographic Use Policy. When the Meta Model API opened in public beta in July, it was available to US developers only — check Meta's policy for your country.

Sources

Sources & change log

Change log

  1. Page published. Muse Wonder (Preview) prices checked against Meta’s pricing and models pages — price version 2026-10-14.1.

Pilot

Already paying an LLM API bill?

Join the pilot for project cost reviews and budgets. Two questions first — we only ask for your email if it fits.

Pilot · Step 1 of 2

Join the pilot

We’re testing project cost reviews with developers who already pay an LLM API bill. Two quick questions first.

Do you have an AI app live in production or in steady trial?

Do you pay an LLM API bill every month?

We only ask for your email if the pilot fits you.