fix(rime): remove Arcana model support

[naszzz]

Summary

• Remove Arcana from the Rime plugin runtime paths, public model types, examples, and tests.
• Change the no-argument Rime TTS model to Coda while keeping the existing Luna speaker.
• Keep Lyra for callers that explicitly select Coda without a speaker.
• Apply repetition_penalty, temperature, top_p, max_tokens, and other supported request controls to Coda.
• Add a release changeset with the migration deadline.

⚠️ Rime's public cloud cutoff is August 15, 2026 at 12:00 UTC. Set modelId to coda when you upgrade. Luna is available on Coda, so the no-argument path keeps its established speaker for voice continuity.

Migration notice: https://docs.rime.ai/docs/arcana-sunset

[changeset-bot]

🦋 Changeset detected

Latest commit: 1e6af9c

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 39 packages

Name	Type
@livekit/agents	Major
@livekit/agents-plugin-rime	Major
@livekit/agents-plugin-anam	Major
@livekit/agents-plugin-anthropic	Major
@livekit/agents-plugin-assemblyai	Major
@livekit/agents-plugin-azure	Major
@livekit/agents-plugin-baseten	Major
@livekit/agents-plugin-bey	Major
@livekit/agents-plugin-cartesia	Major
@livekit/agents-plugin-cerebras	Major
@livekit/agents-plugin-deepgram	Major
@livekit/agents-plugin-did	Major
@livekit/agents-plugin-elevenlabs	Major
@livekit/agents-plugin-fishaudio	Major
@livekit/agents-plugin-google	Major
@livekit/agents-plugin-hedra	Major
@livekit/agents-plugin-hume	Major
@livekit/agents-plugin-inworld	Major
@livekit/agents-plugin-krisp	Major
@livekit/agents-plugin-lemonslice	Major
@livekit/agents-plugin-liveavatar	Major
@livekit/agents-plugin-livekit	Major
@livekit/agents-plugin-minimax	Major
@livekit/agents-plugin-mistral	Major
@livekit/agents-plugin-mistralai	Major
@livekit/agents-plugin-neuphonic	Major
@livekit/agents-plugin-openai	Major
@livekit/agents-plugin-perplexity	Major
@livekit/agents-plugin-phonic	Major
@livekit/agents-plugin-protoface	Major
@livekit/agents-plugin-resemble	Major
@livekit/agents-plugin-runway	Major
@livekit/agents-plugin-sarvam	Major
@livekit/agents-plugin-silero	Major
@livekit/agents-plugin-soniox	Major
@livekit/agents-plugin-tavus	Major
@livekit/agents-plugins-test	Major
@livekit/agents-plugin-trugen	Major
@livekit/agents-plugin-xai	Major

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[devin-ai-integration]

Devin Review found 1 new potential issue.

View 4 additional findings in Devin Review.

[devin-ai-integration]

🟡 Voice tuning settings are silently ignored when the older Rime voice model is still selected

Tuning values such as speed, randomness and length limits are dropped from the request (modelParams at plugins/rime/src/tts.ts:92-109) whenever the older arcana model name is passed, so speech is generated with default settings instead of the configured ones and no error is raised.
Impact: Users who still specify the old Rime model get audio that ignores their configured voice tuning, with no warning.

Parameter branch no longer matches the legacy model id

modelId is typed TTSModels | string (plugins/rime/src/tts.ts:56), so 'arcana' is still accepted at runtime even though it was removed from the TTSModels union (plugins/rime/src/models.ts:6). Previously modelParams had an opts.modelId === 'arcana' branch that forwarded repetition_penalty, temperature, top_p, max_tokens and timeScaleFactor. After this change only 'coda' and *mist* are matched, so an arcana request falls through and emits only lang. Those keys are also explicitly excluded from the passthrough loop in fetchPayload (plugins/rime/src/tts.ts:139-144), so they cannot reach the payload by any other route. The result is a silently under-configured request rather than an explicit error, which contradicts the stated intent of keeping the legacy runtime path working during the compatibility window. Either restore the legacy branch or reject the legacy model id with a clear error.

Prompt for agents

In plugins/rime/src/tts.ts, modelParams() previously had a dedicated branch for modelId === 'arcana' that forwarded repetition_penalty, temperature, top_p, max_tokens and timeScaleFactor. That branch was removed and replaced by a 'coda' branch, but modelId is typed `TTSModels | string`, so callers can still pass 'arcana' at runtime. Because those keys are also on the exclusion list in fetchPayload(), an 'arcana' request now silently loses all of its tuning parameters. Decide on an explicit behavior: either keep forwarding the same parameter set for 'arcana' during the compatibility window, or fail fast in resolveOptions() with a clear error telling the caller to switch to 'coda'.

---

Was this helpful? React with 👍 or 👎 to provide feedback.

[naszzz]

Intentional under the revised scope. This PR now removes Arcana runtime support and directs users to Coda.