Meta · Community reports
Jev + Muse Wonder: a typed judge inside Muse and Muse Code
Meta released Muse Wonder on 12 October 2026. The cited Muse workflows install Jev as a typed judge inside Muse or Muse Code. Those posts and READMEs date from 19 September to 1 October 2026 and name Muse or Muse Code.
Compiled by jevmodel.ai. We read the cited posts and documentation. The measurements stay with their authors. We have not rerun them.
Muse Wonder, Muse Code, and Muse are three products
Meta’s 12 October 2026 model page lists Muse Wonder (muse-wonder-1.0) on the Meta Model API and in Muse Code at $4 input, $0.20 cached input, and $20 output per million tokens, with a 1M-token context window. Muse Wonder Full (muse-wonder-full-1.0) is on Muse Code’s Power Usage plan at $50 a month and Maximum Usage at $200 a month. The page prints no public per-token API price for Full.
The research blog the same day says Muse Wonder will power experiences in the Muse assistant, first for subscribers. Muse is that assistant. Muse Code is the terminal coding agent. Muse Wonder is the language model those products can run.
TypeSafe’s coding-agents page lists Muse Spark among coding-agent LLMs that Jev does not replace. There is no model: "jev-latest" setting that turns a coding agent into a Jev-powered agent. Jev returns a choice, a score, or a noul. Muse Wonder writes text. The page names Muse Spark. Muse Wonder is Meta’s later coding and agent model.
Sources:[160] Meta[161] Meta Superintelligence Labs[162] TypeSafe AI[1] TypeSafe AI
Skills that teach Muse and Muse Code to call Jev
aaddrick’s building-with-typesafe-jev README at commit a46856c, dated 1 October 2026, says the unofficial skill installs in Claude Code, Codex, Antigravity CLI, Muse, and Muse Code. For Muse it copies the skill folder into ~/workspace/skills/ and rewrites the SKILL.md header. For Muse Code it runs muse skills install with --scope user. The skill is Markdown. A TypeSafe key is optional and lets the agent check field names against the live API.
TypeSafe publishes its own skill. The 15 October 2026 agent-skill page names the Claude Code plugin from typesafe-ai/skills and npx skills add. It does not name Muse or Muse Code in those installers. aaddrick’s README says the two skills have different names and do not conflict. The evals in that README used claude-sonnet-5 writing Jev code, with no TypeSafe key, so graders read the generated code. They are not a Muse Wonder runtime score.
theglitcharchitect/muse-skills at commit f0cc9cc installs a Jev decision layer plus 23 skills into ~/workspace/skills/. The bundled demo runs against a mock gateway with no key. Live mode uses VERCEL_AI_GATEWAY_KEY. Shadow mode logs the verdict and still runs the action. Spending, sending, deleting, and publishing escalate to a person in both modes. The demo probabilities, including p = 0.03 on a mock rm -rf, come from that mock walkthrough.
Sources:[169] aaddrick[163] TypeSafe AI[170] theglitcharchitect
A Jev router in front of Muse work
On 22 September 2026, @0xCodila posted a Jev + Muse setup that clones Bodila51/muse-jev-playbook. The loop is prompt, then Muse, then a Jev decision, then Muse execution. The post keeps shadow mode for 20 to 50 real decisions, then active mode with act at 0.80 and above, advisory between 0.50 and 0.79, and a human below 0.50. The “100% of my life” and “200x” lines have no measurement next to them.
The README at commit 70edf68 says the playbook is adapted from grok-bot-jev and generalized for Muse/Hatch, custom harnesses, or scripts. Track B says a Muse/Hatch agent already has the credential and follows skill/jev-decision-layer.SKILL.md. The flights case in the same commit is one local run on 22 September 2026 with jev-1.13.0. The agent collected Google Flights prices. Jev ranked months and days. On a 92-day one-way table, month confidence is 1.0 for October, day 0.67 for 5 October, and book-now 0.82. On the round-trip table both trip lengths pick April 2027 at 1.0, and book-now is 0.79, which the file treats as surface rather than auto-act. The file says Jev did not fetch the fares, the numbers are a snapshot, and the run is a method template.
A CRM skill install, and a connector application
On 30 September 2026, @jkpelaez wrote that installing a Jev skill on Muse sent “same person?” and entity-type questions to Jev. He reports 1,300+ CRM records processed, three duplicates merged, and every record typed against his vocabulary. The agent routes on that typed result. The post names Muse, Meta, and a CLI call. It does not name a Muse model id, a Jev version, or a timed comparison.
On 19 September 2026, @phanindra_ai wrote that he had applied for a Muse connector for jev.magicteams.ai, so a Muse could build the Jev. The post is an application. It does not show an approved connector or a runtime log.
Intelligence Index scores are Muse Wonder against other LLMs
Artificial Analysis scored Muse Wonder (max) at 52 and Muse Wonder Full (max) at 56 on Intelligence Index v4.3, against 48 for Muse Spark 1.3 (max). Wonder is on Meta’s first-party API. Full is only for Muse Code users in that post. Wonder token prices in the post match Meta’s $4 / $0.20 / $20. Full has no per-token price there.
Those scores compare Meta’s language models with other language models. They are a different test from Artificial Analysis’s Decision Index for Jev. The playbook, the skills, and the CRM note all predate 12 October 2026 and name Muse or Muse Code.
Sources:[164] Artificial Analysis[165] codila[169] aaddrick[168] jkpelaez
Questions about Jev + Muse Wonder
Is Jev a drop-in replacement for Muse Wonder or Muse Code?
No. TypeSafe’s coding-agents page says Jev is not a drop-in replacement for the LLM behind Muse Spark and similar coding agents. Jev returns a choice, a score, or a noul. Muse Wonder writes text. The cited skills teach Muse or Muse Code to call Jev. They do not swap the Muse Wonder weights.
Sources:[162] TypeSafe AI[169] aaddrick
Did the cited Muse workflows use muse-wonder-1.0?
The playbook post is dated 22 September 2026, the CRM note 30 September, and the aaddrick README commit 1 October. Meta released Muse Wonder on 12 October 2026. Those sources name Muse or Muse Code.
Sources:[165] codila[168] jkpelaez[169] aaddrick[161] Meta Superintelligence Labs
Where does Muse Wonder Full run?
Meta’s model page lists muse-wonder-full-1.0 on Muse Code’s Power Usage plan at $50 a month and Maximum Usage at $200 a month. It prints no public per-token API price. muse-wonder-1.0 is the Model API model at $4 / $20 per million tokens.
Sources:[160] Meta
Is there a scored Jev + Muse Wonder benchmark?
Artificial Analysis’s 52 and 56 are Muse Wonder and Muse Wonder Full against other language models on Intelligence Index v4.3. The flights case is one local jev-1.13.0 run on fare tables collected by an agent, dated 22 September 2026. The sources opened for this report have no scored pairing of hosted Jev with muse-wonder-1.0.
Sources and reading notes
The summaries below are paraphrases. Each link opens the original source, including its surrounding context.
Meta
Source checked:
[160] Muse Wonder model page: variants, price, and Muse Code access
The page lists Muse Wonder (muse-wonder-1.0) on the Meta Model API and Muse Code at $4 input, $0.20 cached input, and $20 output per million tokens, with a 1M-token context window. Muse Wonder Full (muse-wonder-full-1.0) is listed for Muse Code Power Usage ($50/month) and Maximum Usage ($200/month), with no public per-token API price.
Official product page as accessed 15 October 2026. The benchmark table on the same page compares Muse Wonder and Muse Wonder Full with Claude Fable 5.1, GPT-6 Astra, and Claude Opus 5.5. The page does not mention Jev. Jev is listed here because the note is cited from the Jev and Muse Wonder report.
Meta Superintelligence Labs
Source checked:
[161] Muse Wonder research blog, 12 October 2026
The 12 October 2026 post introduces Muse Wonder and Muse Wonder Full for personal and professional agent work. Muse Wonder is available through the Meta Model API and Muse Code. Muse Wonder Full is available for Muse Code subscribers. The post says Muse Wonder will power experiences in the Muse assistant, first for subscribers.
Vendor research post. Active Memory and Active Focus are Meta’s names for context carry-over and sustained effort. The post does not mention Jev. Jev is listed here because the note is cited from the Jev and Muse Wonder report.
TypeSafe AI
Source checked:
[162] Jev is not a drop-in for Muse Spark or other coding-agent LLMs
TypeSafe says Jev is not a drop-in replacement for the LLM behind Claude Code, Cursor, opencode, Copilot, Muse Spark, Grok Bot, or similar tools. There is no model: "jev-latest" setting that turns a coding agent into a Jev-powered agent. The page tells readers to keep the coding-agent LLM and use Jev separately for typed decisions.
Official coding-agents page as accessed 15 October 2026. It names Muse Spark, not Muse Wonder. Muse Wonder is Meta’s later coding and agent model.
TypeSafe AI
Source checked:
[1] Jev and the Choice, Score and Noul primitives
The documentation describes Jev as a System One model that evaluates state and typed questions. Choice selects an option, Score evaluates a rubric, and Noul returns a probability for a statement.
API behavior documented by the provider. The page describes the three question types and includes no latency, accuracy, or cost comparison.
aaddrick
Source checked:
[169] Unofficial Jev skill with Muse and Muse Code installers
The README at commit a46856c, dated 1 October 2026, says the unofficial skill installs in Claude Code, Codex, Antigravity CLI, Muse, and Muse Code. Muse copies the skill into ~/workspace/skills/. Muse Code uses muse skills install with --scope user. The skill is Markdown. A TypeSafe key is optional.
Community skill, not TypeSafe’s. The snapshot date in the file is 25 September 2026. The evals in the same README used claude-sonnet-5 writing Jev code, not a Muse Wonder runtime. The commit predates the 12 October Muse Wonder release.
TypeSafe AI
Source checked:
[163] TypeSafe agent skill for coding agents
The official skill gives a coding agent the TypeSafe API types, patterns, and evaluation structure. Installers named on the page are the Claude Code plugin from typesafe-ai/skills and npx skills add typesafe-ai/skills --skill typesafe-ai.
Official skill page. It does not name Muse or Muse Code in the installers shown on 15 October 2026.
theglitcharchitect
Source checked:
[170] muse-skills: Jev router plus 23 skills for Muse agents
Commit f0cc9cc installs a Jev decision layer and 23 skills into ~/workspace/skills/. The bundled demo runs against a mock gateway with no key. Live mode uses VERCEL_AI_GATEWAY_KEY. Shadow mode logs the verdict and still runs the action. Spending, sending, deleting, and publishing escalate to a person in both modes.
Commit dated 21 September 2026. The demo probabilities (including p = 0.03 on a mock rm -rf) are from that mock walkthrough. The README names Muse agents and does not name muse-wonder-1.0.
codila (@0xCodila)
Published:
Source checked:
[165] Jev + Muse playbook post with a seven-step setup
The post describes a prompt → Muse → Jev decision → Muse execution loop and links Bodila51/muse-jev-playbook. Shadow mode for 20–50 decisions, then active mode with act at 0.80 and above, advisory between 0.50 and 0.79, and human escalation below 0.50.
Personal setup post dated 22 September 2026, three weeks before Muse Wonder. The “100% of my life” and “200x” lines have no measurement next to them. The post names Muse, not muse-wonder-1.0.
codila
Source checked:
[166] muse-jev-playbook README at commit 70edf68
The README adapts grok-bot-jev for Muse/Hatch, custom harnesses, or scripts. Track A uses the TypeSafe SDK. Track B says a Muse/Hatch agent already has the credential and follows skill/jev-decision-layer.SKILL.md. Policy bands are act at 0.80 and above, surface between 0.50 and 0.79, and escalate below 0.50.
Single commit dated 22 September 2026. The file says the A/B numbers in examples/ are one local run, a method template rather than a benchmark.
codila
Source checked:
[167] Flight-date case: jev-1.13.0 on one local fare table
One local run on 22 September 2026. The agent collected Google Flights prices; jev-1.13.0 ranked months and days. On a 92-day one-way table, month confidence is 1.0 for October, day 0.67 for 5 October, and book-now 0.82. On the round-trip table, both trip lengths pick April 2027 at 1.0, and book-now is 0.79.
The file says Jev did not fetch the fares, the numbers are a snapshot, and the run is not a benchmark. Fares need re-checking before a purchase.
TypeSafe AI
Source checked:
[2] Confidence, probabilities and fallback decisions
Choice and Score include confidence derived from their probability distributions. Noul has no separate confidence field. TypeSafe recommends choosing thresholds for the domain and consequences of an action.
Implementation reference. A high confidence value can still sit on a wrong decision.
jkpelaez (@jkpelaez)
Published:
Source checked:
[168] Jev skill on Muse for CRM dedup and entity typing
The author says installing a Jev skill on Muse sent “same person?” and entity-type questions to Jev. He reports 1,300+ CRM records processed, three duplicates merged, and every record typed against his vocabulary. The agent routes on the typed result.
First-hand field note dated 30 September 2026. The post names Muse and a CLI call. It does not name a Muse model id, a Jev version, a timed comparison, or Muse Wonder.
Phanindra Reddy (@phanindra_ai)
Published:
Source checked:
[171] Application for a Muse connector on jev.magicteams.ai
The author wrote that he had applied for a Muse connector for jev.magicteams.ai, so a Muse could build the Jev.
An application dated 19 September 2026. The post does not show an approved connector, a runtime log, or a Muse model id.
Artificial Analysis (@ArtificialAnlys)
Published:
Source checked:
[164] Artificial Analysis scores Muse Wonder 52 and Wonder Full 56
Artificial Analysis scores Muse Wonder (max) at 52 and Muse Wonder Full (max) at 56 on Intelligence Index v4.3, against 48 for Muse Spark 1.3 (max). Muse Wonder is on Meta’s first-party API. Muse Wonder Full is only for Muse Code users. Wonder token prices in the post are $4 input, $0.20 cached, and $20 output per million tokens. Full has no per-token price in the post.
Independent language-model index. These scores compare Meta’s models with other language models. They are a different test from Artificial Analysis’s Decision Index for Jev. Jev is listed here because the note is cited from the Jev and Muse Wonder report.