New Describe a workflow. Raq.com builds it for you and runs it on autopilot. Try Automations free →
Multi AI Chat AI Team Comms Photoshoot AI Task Board Top 20 AI Consensus AI AI Image Lab AI Marketing Campaigns AI Time Tracker Voice Calls AI Social Scheduler AI Video Lab AI Automations AI Ask My Agent AI Orchestrator Custom Tools AI AI Decisions AI Web Previews E-Sign Website Build & Host AI Client Portal Help Desk Page Feedback CRM Arbiter AI Design Swarm AI Project Base Ghostwriter AI Pretty Docs AI Secure Send Market Intel AI AI Music Lab AI KPI Hub Site Uptime Page Watch Server Monitor Knowledge AI Render to Reality AI Vector Motion Lab AI AI Benchmarker AI Website Chatbox AI Co-Write File Drop Contract Auditor AI Spreadsheet AI Creative Review Priorities AI Compliance Tracker Anti-slop Writer AI Rewriter AI Brand Guidelines AI Screenshot Studio Design MCP AI Screen Recorder Background Remover AI Image Optimiser SEO & AEO Checker AI Transcribe AI Vocalise AI Email Signatures AI Personas AI Recall 280daily Journal Meal Planner AI Health Insights AI Nature Work Passkey Vault Form Builder AI Prompt Library AI Skill Library AI PDF Page Studio File Converter Diff Viewer Receipt Extractor AI Branded CV AI FinUK Short URL QR Generator OneDrive Search AI
Raq.com
Back to Blog
· The Raq.com Team

Agent Operator HQ Now Supports Muse Code

Agent Operator HQ Now Supports Muse Code

Agent Operator HQ now supports Muse Code, Meta's coding agent, through a built-in session template.

Muse Code stays installed and signed in on your own computer. Agent Operator HQ shows when it's working and when it needs an approval, and sends queued prompts once it's ready. Nature Work does the same on your phone. Reopen a closed session and it picks up the same conversation.

Muse Wonder

Meta released Muse Wonder and a larger Muse Wonder Full on 12 October. Muse Wonder is on every Muse Code plan and on Meta's API at $4 per million input tokens and $20 per million output tokens. Wonder Full is only in Muse Code, on the $50 and $200 plans.

On Artificial Analysis's Coding Agents Index, Muse Code with Wonder Full scored 68, level with Claude Code on Sonnet 5.5. Claude Opus 5.5 and GPT-6 Astra still lead Terminal-Bench 4.0.

Coding Agents Index

Each model tested inside its coding agent

  • Muse Code + Wonder Full 68
  • Claude Code + Sonnet 5.5 68
  • Claude Code + Opus 5.5 66
  • Muse Code + Wonder 63
  • Muse Code + Spark 1.3 54

Terminal-Bench 4.0

Terminal tasks, pass@1

  • Claude Opus 5.5 59.6%
  • GPT-6 Astra 59.1%
  • Muse Wonder Full 57.1%
  • Muse Wonder 52.0%

DeepSWE 1.1

Software engineering tasks, pass@1

  • Muse Wonder Full 79.0%
  • Muse Wonder 78.8%
  • Claude Opus 5.5 74.2%
  • GPT-6 Astra 74.1%
Coding Agents Index and Terminal-Bench 4.0 from Artificial Analysis (12 October 2026). DeepSWE 1.1 from Meta's Muse Wonder model card.

Getting started

Install Muse Code and run muse login on the computer that runs Agent Operator HQ, then start a session from the Muse Code template. It's available on every Raq.com plan.