Muse Wonder
pricing, made clear.
Know what you’ll pay for Meta’s AI model before you build. Compare verified API rates, understand subscription access, and turn your token usage into a practical budget.
Muse Wonder · per 1M tokens
10K input + 2K output tokens. No caching or search.
Check the official price table ↗
One model. Three token rates.
Your API bill depends on the tokens sent, the portion served from cache, and the output generated. These are separate rates, not three subscription tiers.
Input tokens
New content sent to the model, including uncached prompts and conversation.
Cached input
An eligible repeated prefix costs 95% less when it is actually served from cache.
Output tokens
Budget using the provider’s billable output count, not only the visible answer.
Meta pricing and rate limits ↗. Model ID: muse-wonder-1.0. No separate long-context premium is listed.
Wonder Full is offered through eligible Muse Code subscriptions. A subscription does not cover the additional API keys used by your own app. See how access and billing differ.
What could your workload cost?
These are illustrative calculations, not measured bills. Token counts are chosen assumptions; actual application usage will vary.
| Illustrative workload | Input / output per request | Cache assumption | Total estimate |
|---|---|---|---|
| Short chat · 1,000 requests | 2,000 / 500 tokens | 0% | $18.00 |
| Document review · 100 requests | 100,000 / 10,000 tokens | 50% | $41.00 |
| Agent workflow · 10,000 requests | 10,000 / 2,000 tokens | 80% + 1 search/request | $521.00 |
The agent example includes 10,000 search queries ($25). Totals exclude taxes, other tools and reseller charges.
Calculate your Muse Wonder API budget
Enter total input and billable output per request, then choose your cache assumption and the number of requests in your budget period. Search queries are added separately. The calculation runs locally and needs no API key.
Set your workload
Your Muse Wonder estimate
Total for your chosen period$0.08 per request · Standard mode · USD
$21.00 — a cost comparison, not a capability comparison.
Estimate only. Excludes taxes, currency conversion, reseller fees, other tools and variation in actual usage. No API key or personal data required.
Wonder and Wonder Full
have different billing routes.
Regular Wonder has published API rates. Full is a separate model available by subscription in Muse Code.
| Model | Meta Model API | Muse Code |
|---|---|---|
| Muse Wonder | Pay as you go · Standard mode | Pay as you go, plus all subscription plans |
| Muse Wonder Full | Not offered on Model API | Power Usage and Maximum Usage |
Meta model availability ↗ · Muse Code subscriptions ↗. Benefits and checkout availability can vary by region.
Make an informed choice.
Understand every line item.
Input, cached input, output and search charges, with worked examples.
Read the API guide →Wonder or Wonder Full?
Which plans unlock each model and how subscription billing differs.
Compare plans and access →Compare the same workload.
Wonder and Spark costs, without invented performance claims.
Compare model costs →
The model behind the pricing.
Meta introduced Wonder and Wonder Full on 12 October 2026 for complex work and long-running agents. We focus on what you can verify and calculate: rates, access rules and billing mechanics.
No invented benchmarks, customer reviews or claimed hands-on tests.
Read our sources and calculation notes →Muse Wonder pricing FAQ
How much does Muse Wonder cost?
Meta Model API lists Standard mode at $4 per million uncached input tokens, $0.20 per million cached input tokens and $20 per million output tokens. Search queries cost extra. These are usage rates, not a monthly plan. Official source ↗
Is Muse Wonder Full available through the API?
Meta lists Full as a Muse Code subscription model on Power Usage and Maximum Usage. It does not list a Model API token rate. Raising reasoning effort on regular Wonder does not select Full. Official source ↗
Does a Muse Code subscription cover my app’s API calls?
The subscription applies to Muse Code’s connected CLI credential. Additional API keys use pay-as-you-go billing. Budget these separately when running your own app or agent. Official source ↗
Does Wonder have an API Contributor discount?
Wonder has no Contributor variant on Model API. Eligible Muse Code subscriptions have a separate Contributor setting, providing more usage in exchange for allowing content to be used for model training. Official source ↗
Does a long context cost extra?
Meta states the same token rates apply at short and long context lengths. Additional tokens still increase your total cost, even without a separate long-context premium. Official source ↗
How much does web search add?
Meta lists $2.50 per 1,000 search queries on top of token charges: $0.0025 per query. Count actual tool queries rather than assuming one query per prompt. Official source ↗
Is the Muse consumer app the same purchase as Muse Code?
Muse Code is a developer coding product. Consumer app subscriptions are a separate purchase context. This guide does not claim an app plan unlocks Full or includes general API credit. Check the relevant official checkout. Official source ↗
Are the examples actual bills?
They are arithmetic estimates from published rates and explicitly chosen workloads, not measured invoices, hands-on tests or claims about the tokens a task will need.
Find your next Muse guide.
Choose the right model, connect the API, install Code, check Full access or understand open weights. Our source-linked guides separate product facts from estimates.