Meta Muse Pricing: Free, Power, and Maximum Plans

Meta Muse pricing starts at zero dollars for a limited free tier, moves to $20 per month for the Power plan with 500 million tokens per week, and tops out at $100 per month for the Maximum plan with 3 billion tokens per week.

At Layer3Labs, we track agent compute costs across real deployment environments to help teams evaluate automation budgets before committing capital.

The consumer personal Artificial Intelligence (AI) agent covered here is what Meta announced on September 8, 2026, accessible on Meta's official Muse portal. This product is distinct from the Muse Spark foundational model family, the Muse Code command-line interface developer tool, and earlier image generators.

Every tier operates on a recurring weekly compute quota. We outline the full rate structure, explain the billing rules from Meta's Help Center, and clarify what those token limits actually mean for everyday use.

The Three Meta Muse Subscription Tiers

Meta Muse offers three service tiers that scale compute quotas from basic conversational interactions to intensive multi-agent background tasks.

Meta's announcement post describes Meta Muse as an autonomous personal assistant built to handle planning, cross-app execution, and purchases. The service runs on a dedicated Muse Secure Virtual Machine (VM) to protect credentials while executing tasks. Access currently spans the United States (US) across mobile applications, web browsers, and messaging channels.

The entry point is a free tier that gives users access to core agent capabilities with a periodic usage limit. Meta does not publish the exact numeric token cap for free accounts, but the allowance refreshes at set intervals. Once that quota depletes, users must wait for the reset or upgrade to a paid subscription.

Paid access splits into two distinct tiers: Power and Maximum. The Power tier costs $20 per month and supplies 500 million Muse tokens each week. The Maximum tier costs $100 per month and provides 3 billion Muse tokens weekly. Neither tier requires an annual commitment.

• Free tier: zero dollars per month with a recurring periodic token quota that resets automatically.
• Power plan: $20 per month for 500 million Muse tokens delivered on a weekly schedule.
• Maximum plan: $100 per month for 3 billion Muse tokens delivered on a weekly schedule.
• Infrastructure: all tiers run on the isolated Muse Secure VM architecture.

Weekly Token Allowances and the Task Conversion Gap

Meta measures subscription allowances strictly in weekly Muse tokens rather than completed real-world tasks.

The Power plan provides 500 million tokens per week, while the Maximum plan provides 3 billion tokens per week. These large numbers reflect internal reasoning, tool calls, and background state monitoring handled by the underlying Muse Spark model. Because autonomous agents generate real prompt overhead while checking web pages, reading documents, and maintaining session state, token burn accelerates quickly during complex tasks.

Meta has not published how many completed real-world actions a given token allowance buys. Users cannot look up an official conversion rate showing how many travel itineraries, form submissions, or email drafts correspond to 500 million tokens. That conversion rate is the single missing metric in Meta's subscription documentation.

Without published task conversions, subscribers must gauge usage based on workflow intensity. Routine drafting and quick lookups consume modest tokens, whereas leaving Meta Muse running continuous background research on flight prices or multi-step shopping tasks drains the weekly pool much faster.

• Weekly reset: token allocations arrive every seven days rather than pooling as an annual or monthly lump sum.
• No rollover: unused weekly tokens expire when the next cycle begins.
• Missing conversion rate: Meta does not disclose the average token cost of booking travel, filling forms, or managing calendars.
• Background overhead: autonomous monitoring across connected applications consumes tokens continuously even when the user closes the interface.

Differences Between the Power and Maximum Plans

The primary difference between the Power and Maximum plans is a sixfold increase in weekly token volume for a fivefold increase in monthly price.

Choosing between Power at $20 per month and Maximum at $100 per month comes down to the frequency of background autonomous execution. The Power plan suits users who invoke Meta Muse interactively throughout the day for focused requests like drafting correspondence, organizing schedules, or compiling research summaries. At 500 million tokens per week, occasional background work remains viable, but high-frequency polling can exhaust the quota prematurely.

The Maximum plan targets users who rely on Meta Muse to run persistent background monitoring and complex multi-application workflows. At 3 billion tokens weekly, the Maximum tier accommodates constant agent activity, including repetitive web verification, continuous calendar adjustments, and automated multi-step transactions.

Both plans include identical access channels across web, mobile, and WhatsApp. Neither plan grants priority processing speed or exclusive features in the current release. The pricing ladder simply buys higher volume for intensive agentic routines.

• Cost comparison: Power costs $20 monthly, while Maximum costs $100 monthly.
• Volume comparison: Maximum delivers 3 billion tokens weekly compared to 500 million on Power.
• Efficiency ratio: Maximum provides six times the token volume for five times the subscription price.
• Functional parity: both tiers share the same access channels and security protocols without feature gating.

Billing Mechanics and Account Eligibility Rules

Meta Muse subscriptions renew automatically each month on the original calendar date of purchase until canceled.

According to Meta's billing terms, subscriptions process recurring payments automatically through the payment method tied to the user's account. If a user subscribes on the fourteenth day of a month, subsequent renewals charge on the fourteenth of each subsequent month. Canceling prevents future charges while preserving access through the remainder of the active billing period.

Account eligibility carries specific geographic and age constraints. Meta limits current subscriptions to users aged 18 and older located in supported regions. The initial rollout remains restricted to the United States under a limited testing framework, meaning access is not universal across all international accounts.

Payments for subscriptions remain distinct from in-agent purchases. When Meta Muse buys retail products on behalf of a user, it uses Stripe Link checkout rails with single-use virtual card numbers rather than drawing from the monthly subscription balance. Subscription fees cover compute capacity alone.

• Renewal schedule: recurring monthly billing occurs on the calendar day of original enrollment.
• Cancellation policy: access continues until the current monthly cycle concludes after cancellation.
• Age restriction: subscribers must be at least 18 years old to create a paid account.
• Regional availability: restricted to the United States during the limited public testing phase.

Disambiguating the Meta Muse Product Line

Meta Muse is a consumer-facing personal task agent, separate from Meta's developer tools and underlying foundation models.

The name Muse appears across several distinct technologies within Meta and across the broader software industry. Within Meta, the personal agent detailed on Meta's Muse product page relies on the Muse Spark foundation model. Muse Spark is the raw language model trained for agentic reasoning. It is separate from the consumer application that end users purchase through app stores.

Similarly, Meta operates Muse Code, a dedicated command-line interface agent tailored specifically for software engineering tasks. Muse Code targets technical developers writing code in terminal environments, whereas Meta Muse serves general personal productivity like planning travel, handling inbox tasks, and coordinating errands. Meta also previously maintained Muse Glimmer as a separate internal model, as well as an older, separate image-generation project that also carried the Muse name.

Outside Meta, independent software developers market tools under similar titles, including companion applications and video generation suites. Confirming that a subscription applies to Meta Muse ensures you receive the personal assistant running on Meta Secure VM rather than an unrelated third-party utility.

• Meta Muse: the autonomous personal assistant operating on mobile, web, and WhatsApp.
• Muse Spark: the foundational reasoning model powering the personal agent.
• Muse Code: a distinct developer CLI tool focused exclusively on programming tasks.
• Third-party products: independent software products using the phrase Muse AI have no affiliation with Meta.

Workloads Meta Muse Does Not Serve

Meta Muse does not support enterprise multi-seat administration, programmatic API pipelines, or unrestricted merchant purchases.

Organizations seeking shared team workspaces, programmatic endpoints, or centralized enterprise compliance cannot use Meta Muse. Meta Muse serves individual personal productivity. It does not provide enterprise administration or programmatic software integration. For programmatic agent pipelines or custom software integration, teams must use direct developer APIs rather than consumer subscriptions.

Users expecting universal retail automation will also encounter friction. While Meta Muse integrates Stripe Link purchase protection for retail transactions, merchant adoption varies widely. On September 20, 2026, roughly twelve days after launch, Amazon began blocking Meta Muse from completing purchases on its marketplace. Autonomous checkout capabilities depend entirely on whether individual vendors allow automated agents on their storefronts.

International teams and individuals under 18 years old are also excluded. The limited US testing release means anyone requiring worldwide availability or youth account support must wait for expanded geographic rollout.

• Enterprise teams: no centralized user management, team billing, or shared corporate workspaces.
• Software developers: no programmatic API keys or custom webhook pipelines for backend integration.
• Universal shoppers: merchants such as Amazon actively block agentic checkout attempts on their platforms.
• Non-US users: accounts remain inaccessible outside supported domestic testing areas.

Factors That Would Change This Cost Assessment

Our assessment of Meta Muse pricing would change if Meta publishes concrete task conversion benchmarks or broadens merchant checkout acceptance.

The value proposition of the $20 Power plan and $100 Maximum plan currently hinges on trial and error. If Meta publishes standardized consumption metrics showing exactly how many web searches, email workflows, or schedule optimizations a 500-million-token allocation supports, buyers could calculate concrete return on investment before subscribing.

A change in retail merchant cooperation would also alter the economic calculation. If major retailers like Amazon reverse restrictions or establish official agentic purchasing protocols, the time saved through automated purchasing would justify the higher monthly fees of the Maximum tier. If more merchants implement bot barriers against automated checkout, the practical utility of Meta Muse narrows primarily to drafting and organization.

Before spending $20 or $100 per month, begin with the free tier to measure how quickly your daily habits approach the recurring quota ceiling. To evaluate Meta Muse pricing against your workflow needs, test your standard repetitive tasks on the free allocation before upgrading to a paid subscription.

• Published benchmarks: transparent token consumption rates for common tasks would make cost planning predictable.
• Merchant access: broad acceptance by major e-commerce platforms would elevate the value of paid subscriptions.
• Geographic expansion: international availability would make Meta Muse viable for distributed teams.
• API availability: developer endpoints would turn Meta Muse into a programmable system.