Troubleshoot Muse Code access, billing, and setup

Model missing or access denied

Start by identifying the exact model and path you are using. Muse Code plan access, consumer-app access, and Model API access are different things.

1. Are you asking for a Model API model such as muse-wonder-1.0? Use the documented API path and billing setup instead of treating it as a Muse Code entitlement.
2. Are you using Muse Code? Confirm that the signed-in Meta Model API account is the account that holds the intended subscription, then check the models presented in the supported Muse Code experience.
3. If the expected model is still absent, review the current official subscription plan details and account eligibility. Availability can depend on the plan, account, and region.
4. If the UI lists the model but a request is denied, preserve the exact error and timestamp. A denied request can reflect provider entitlement, billing, policy, or service state that local changes cannot resolve.

Unexpected charges or quota

If you see a charge you did not expect

1. Identify whether it appears in subscription history or Model API usage. They are separate billing paths.
2. For subscription billing, review the plan, renewal date, receipts, and payment method in the official subscription management flow.
3. For API billing, inspect the account, API key, model ID, request volume, token usage, and tool usage for the relevant time range.
4. Include web-search usage in an API review: the published rate is $2.50 per 1,000 queries, separate from token charges.
5. Compare your findings with the pricing guide and the provider's current billing records. Ask the provider for account-specific clarification when those records do not explain the charge.

If you hit a quota or rate limit

1. In Muse Code, review plan usage and the refresh information shown in the official account experience. The provider documents that a plan limit can require waiting for refresh or changing to an eligible higher plan.
2. In the Model API, distinguish account budget issues from a request rate limit. Rate limits apply per team rather than per API key.
3. For API 429 responses, follow the provider guidance: respect Retry-After when present and use backoff with jitter. Do not fan out retries from several clients at once.

Authentication or configuration

1. Confirm which product you are configuring: supported Muse Code onboarding or a Model API client integration. Do not copy a credential from one path into the other unless the official documentation explicitly supports that use.
2. Reopen the official documentation for the current sign-in or API setup steps. Configuration details and supported clients can change.
3. Check for a local stale session, an expired credential, a changed project directory, or a configuration value pointing at a different account. Record what changed before rotating or removing anything.
4. Never paste an API key, authorization header, browser cookie, or complete environment file into a ticket, screenshot, log, or chat.

For supported setup instructions, use the official Muse Code documentation and the Model API quickstart.

Local state or provider state?

Prepare support evidence

If the official account state and local checks still disagree, prepare a short, redactable support note. It helps support distinguish an entitlement issue from a configuration issue without exposing secrets.

Product path: Muse Code or Model API
App / CLI version: [version]
Model ID or model label: [model]
Billing path: subscription or pay-as-you-go
Account context: [non-secret account identifier, if requested]
Time and timezone: [timestamp]
Exact error or status: [redacted error]
What changed: [brief description]
What you already checked: [brief list]
Secrets omitted: API keys, tokens, cookies, and authorization headers

Include only the fields relevant to the case and follow the provider's support instructions. Support can investigate account-specific state, but this guide cannot promise access restoration, a billing outcome, or a provider-side repair.