Repository navigation
Add the Meta Muse Wonder model card and metadata - #216
Merged
Merged
Conversation
Meta released Muse Wonder on 2026-10-12 (TechCrunch, VentureBeat). Its API id on Meta's Model API is muse-wonder-1.0. PolicyBench holds no Meta key, and OpenRouter's only endpoint for meta/muse-wonder is Meta itself (provider "Meta", serving meta/muse-wonder-20261012), so the row runs through OpenRouter as Kimi, GLM, Qwen, MiniMax and Inkling do. Priced at Meta's list rate, $4 / $20 per 1M input/output tokens with $0.20 cached input and no long-context tier (Meta's launch post and the OpenRouter listing, read 2026-10-12). The serving card waits on the onboarding probes (forced tool vs auto reasoning, gauntlet, smoke). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
4 of 8 tasks
…ses a forced tool Meta's API, which OpenRouter's one endpoint for meta/muse-wonder passes requests to, accepts tool_choice "auto" only: a named-function or "required" choice returns 400. The board's rule is a forced answer tool unless the provider rejects it, so the row runs the JSON contract whole-scenario, as Opus 5.5 and Sonnet 5.5 do, reasoning at the provider default (a one-question probe used 1,913 to 3,089 reasoning tokens with no tools and 2,189 to 2,635 with tool_choice auto). With no forced tool there is no tool_choice auto sensitivity run. Gauntlet (2026-10-13): without the reasoning budget the 3-variable JSON probe stopped at the token limit; with it, 3/3 and 16/16 parsed (2,402 and 2,736 completion tokens, 15s and 18s), about $0.059 per scenario. A two-scenario smoke on the run path completed 2/2 with no missing prediction or explanation for $0.12 (Meta's reported cost equals the reconstruction from list prices). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
MaxGhenis
marked this pull request as ready for review
October 13, 2026 02:50
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…(card review P3) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
MaxGhenis
added a commit
that referenced
this pull request
Oct 13, 2026
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This branch was successfully deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This is the card PR for Meta Muse Wonder, following #201, the Claude Haiku 5.5 card. Its scores join the published board in the next data release, so this PR changes no published artifact.
Facts (2026-10-12)
muse-wonder-1.0, at $4 / $20 per 1M input/output tokens and $0.20 per 1M cached input.meta/muse-wonderis Meta itself (provider "Meta", servingmeta/muse-wonder-20261012), so the row runs through OpenRouter, as Kimi, GLM, Qwen, MiniMax and Inkling do.The card (2026-10-13)
tool_choice: "auto"only. A named-function or"required"choice returns 400 ("only\"auto\"is supported fortool_choice"). The board's rule is a forced answer tool unless the provider rejects it, so the row runs the JSON contract, as Opus 5.5 and Sonnet 5.5 do.tool_choice: "auto". It returns a short reasoning summary and an encrypted trace. With no forced tool there is notool_choiceauto sensitivity run, because that run measures how much a forced tool suppresses reasoning.~/reviews/policybench-muse-wonder/{probes,onboard,smoke}.Plan
The board run started at e6f69d4 on the release branch (#217), which carries this card.
🤖 Generated with Claude Code