Meta AI Releases Muse Code (Beta): A Terminal Coding Agent Powered by the New Muse Spark 1.2 Model
https://www.marktechpost.com/2026/08/05/meta-superintelligence-labs-releases-muse-code/📌 【Meta AI】推出 Muse Code Beta：具備持久性背景代理能力的終端機編程助手
TL;DR：Muse Code 透過 Muse Spark 1.2 模型，實現能在大型專案中進行長期、複雜任務的終端機編程代理。
隨著 AI 代理（Agent）從簡單的對話轉向複雜的軟體工程任務，Meta AI 推出了 Muse Code (Beta) 版本。這不僅僅是一個對話框，而是一個能在大型程式碼庫中規劃、編寫並驗證結果的終端機代理，其核心動力來自於全新的 Muse Spark 1.2 模型。
🧩 透過背景代理與事件日誌，實現長效且穩定的任務執行
與傳統「針對單一任務啟動一個代理」的模式不同，Muse Code 採用了一種全新的設計理念：
- 異步背景代理 (Async Background Agents)：一組專門的背景代理會在整個工作階段（Session）中保持活躍，而非在每個任務後重新啟動。這種設計能避免重複進行資訊收集，並在處理複雜的多步驟任務時，降低延遲並提升操控精準度。
- 可重啟的事件日誌 (Replay-exact Event Log)：系統會將每一次的模型呼叫、工具執行、使用者核准以及程式碼編輯，都記錄在一個本地的「僅限追加」事件日誌中。這讓代理具備「斷點續傳」的能力，即使發生當機，也能從上次停止的地方精確恢復。
🛠️ 內建三大核心指令：從規劃到壓力測試
Muse Code 提供了一套預設的技能指令，協助工程師管理開發流程：
/plan：將任務轉化為一個需要經過核准的執行計畫。/grill：對該計畫進行壓力測試，直到計畫足以支撐複雜邏輯。/goal：朝著指定的最終目標持續執行任務。
📊 Muse Spark 1.2：針對編程工作流深度優化的模型
Muse Spark 1.2 是對前代 1.1 版本的升級，Meta 特別強調該模型與其測試框架（Harness）進行了協同訓練（Co-trained）。
- 效能提升：在程式碼生成、複雜除錯（Debugging）、程式碼庫理解以及端到端開發工作流方面皆有顯著進步。
- 訓練規模：研究團隊大幅提升了編程任務的訓練算力，並擴大了環境的多樣性。
- 評估基準：
- Terminal-Bench 2.1：測試 89 個任務。
- DeepSWE v1.1：涵蓋 91 個專案、5 種語言的 113 個任務。
- Meta Internal Coding Bench：包含 440 個源自內部實際 Pull Request 的任務。
- 對比基準：實驗中對比了 Grok 4.5、Claude Opus 5、GPT-5.6 Terra、Gemini 3.6 Flash 與 Kimi K3。
💡 深度技術案例：GPU Kernel 優化
Meta 展示了一個極端的測試案例：針對 GPU Kernel 進行迭代優化，該過程包含超過 1,000 次工具呼叫，持續時間長達 24 小時。模型會針對基準進行編寫、編譯、分析效能（Profiling）並逐步改進。在 NVIDIA Hopper GPU 上的測試顯示，模型能針對 KDA 與 MLA kernel 進行高度複雜的優化工作。
⚠️ 使用限制與安裝方式
目前 Muse Code 僅提供 Beta 版本，且根據發布資訊，模型似乎是透過託管方式（Hosted dependency）提供，並未釋出可下載的權重（Weights）。
- 支援平臺：macOS 與 Linux。
- 安裝指令：
curl -fsSL https://dev.meta.ai/install.sh | bash
🎯 實務啟示
對於需要處理大規模既有專案（Legacy Codebase）的工程師來說，這種具備「持久狀態」與「背景代理」能力的工具，比單純的程式碼補全更具實戰價值。它更像是一個能理解專案上下文、並能自主進行長期除錯與優化的虛擬夥伴。
🔗 來源
- 標題：Meta AI Releases Muse Code (Beta): A Terminal Coding Agent Powered by the New Muse Spark 1.2 Model
- 作者／機構：Asif Razzaq @ MarkTechPost
- 連結：https://www.marktechpost.com/2026/08/05/meta-superintelligence-labs-releases-muse-code/
#MetaAI #MuseCode #MuseSpark #CodingAgent #SoftwareEngineering #LLM #MachineLearning #GPUOptimization #DevTools #AIResearch
原始資料 MarkTechPost · 收集於 2026-08-06
摘要原文
Meta AI has released Muse Code (in beta), a terminal coding agent in beta, powered by its new Muse Spark 1.2 model . Meta positions the pair as its next step toward the frontier, with larger models on the way. Muse Code targets complex software engineering across large repositories: it plans changes, writes code, and validates the results. A set of async background agents stays alive for the whole session instead of spawning per task. A local append-only event log records every model call, tool run, approval, and edit, which Meta calls replay-exact and restart-safe. Muse Spark 1.2 was co-trained with the harness itself. Meta also published a kernel-optimization case study running 1,000+ tool calls over as long as 24 hours. Yes. Muse Code ships in beta for macOS and Linux via curl -fsSL https://dev.meta.ai/install.sh | bash . Muse Spark 1.2 is available in Muse Code and the Meta Model API , with expanded global access. The launch post does not mention downloadable weights, so treat this as a hosted dependency. Muse Code runs a simple agent loop plus a set of async background agents. These specialized agents remain active throughout each session. They are not spawned for individual tasks, which Meta says avoids redundant information gathering. They carry out next steps and choose when to report back to the main agent. Meta states this persistence reduces latency and steering on difficult, multi-step tasks. Muse Code uses a local event log. Every model call, tool run, approval, and edit is appended to it. Meta calls this single source of truth replay-exact and restart-safe. After a crash, the agent resumes precisely where it stopped, letting long-running tasks survive failures. Three default skills ship with the agent. /plan turns a task into an approval-gated plan. /grill stress-tests that plan until it holds up. /goal works toward successful completion of the specified objective. Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 . Meta reports gains in code generation, complex debugging, codebase understanding, and end-to-end developer workflows. The research team significantly scaled up training compute on coding tasks and expanded environment diversity. The model keeps its strength in other areas, including general agents. Three important training details: Meta’s methodology report is unusually specific. Terminal-Bench 2.1 uses all 89 tasks, pass@1 over five attempts. DeepSWE v1.1 covers 113 tasks across 91 repositories and five languages. Meta Internal Coding Bench holds 440 tasks derived from real internal pull requests. Runs execute in isolated Daytona cloud sandboxes. Comparisons include Grok 4.5, Claude Opus 5, GPT-5.6 Terra, Gemini 3.6 Flash, and Kimi K3, each with its own agent product. Meta notes its harness may not be tuned for third-party models. For reference, Meta’s model page lists Muse Spark 1.1 at 80.0 on Terminal-Bench 2.1. Meta tested iterative GPU kernel optimization over 1,000+ tool calls, running up to 24 hours. The model writes, compiles, profiles, and progressively improves kernels against a provided baseline. Benchmarks covered KDA and MLA kernels on NVIDIA Hopper GPUs. For KDA, the baseline is the FLA Triton implementation, with third-party kernel libraries prohibited. Muse Spark 1.2 paired a chunk-parallel preparation kernel with a sequential inter-chunk scan. For MLA, the reference is PyTorch at batch size 1, 64 heads, sequence length 8192, and latent dimension 512. The model built a two-kernel Triton pipeline that reuses the shared KV latent as both K and V. Check out the Technical details , Model (Muse Spark 1.2) and Evaluation Methodology . Also, feel free to follow us on Twitter and don’t forget to join our 150k+ML SubReddit and Subscribe to our Newsletter . Wait! are you on telegram? now you can join us on telegram as well. Need to partner with us for promoting your GitHub Repo OR Hugging Face Page OR Product Release OR Webinar etc.? Connect with us The post Meta AI Releases Muse Code (Beta): A Terminal Coding Agent Powered by the New Muse Spark 1.2 Model appeared first on MarkTechPost .
由 tencent/hy3:free 自動生成