Checked October 11, 2026 against OpenCode’s official Zen documentation and its model catalog endpoint. Meta’s model page was not reliably retrievable during this review, so we do not repeat unverified benchmark or direct-API pricing claims.
| Zen entry | Model ID | Published endpoint family |
|---|---|---|
| Muse Spark 1.3 | muse-spark-1.3 | Responses: https://opencode.ai/zen/v1/responses |
| Muse Spark 1.3 Contributor Free | muse-spark-1.3-contributor-free | Responses: https://opencode.ai/zen/v1/responses |
Select it without confusing billing routes
- Check the live catalogModel availability and names can change. Confirm the exact entry in your account rather than assuming the standard model is free.
- Connect OpenCode ZenThe published Zen guide uses /connect, followed by provider selection and an API key. Keep credentials out of screenshots, source control and shared configuration.
- Choose the exact variantUse /models to inspect available entries. For the documented V1 configuration convention, the ID is prefixed with opencode/, for example opencode/muse-spark-1.3. Use the matching V2 provider docs if migrating.
- Read privacy and spend settingsReview Contributor participation terms and model-specific data handling before submitting company or personal information. Check credit balance, auto-reload and monthly spend controls.
- Run a bounded evaluationUse a non-sensitive test repository, require approval for writes and inspect output, tests and usage charges. This guide does not claim an authenticated task was benchmarked.
Free software, free model and paid plan are different
OpenCode is the client. Zen is an optional model gateway, and Go is a separate plan with its own conditions. A free entry in the model picker does not make every request or every variant free. Read the current terms for the exact route, especially when Contributor data may be used to improve models.
If the model is missing or requests fail
First compare your configured provider, exact model ID and endpoint family with the official catalog. Check account access and current service status. Do not send a Responses model blindly to a Chat Completions endpoint, or assume another client’s configuration is compatible. Report a redacted error and client version rather than cycling private credentials.
FAQ
Are the standard and Contributor Free variants identical billing choices?
No. They have different IDs and may have different pricing and data-use conditions. Check the live catalog and participation terms.
Can I send private code to the free variant?
Only after reviewing its data-use terms and your organization’s policy. Do not assume free means zero retention or no model training.
Does this prove it outperforms other coding models?
No. This guide verifies documented access and configuration, not comparative performance. Benchmark a representative task with your own quality and cost criteria.