Meta AI has launched Muse Code (in beta), a terminal coding agent in beta, powered by its new Muse Spark 1.2 mannequin. Meta positions the pair as its subsequent step towards the frontier, with bigger fashions on the best way. Muse Code targets advanced software program engineering throughout giant repositories: it plans adjustments, writes code, and validates the outcomes. A set of async background brokers stays alive for the entire session as a substitute of spawning per process. An area append-only occasion log information each mannequin name, instrument run, approval, and edit, which Meta calls replay-exact and restart-safe. Muse Spark 1.2 was co-trained with the harness itself. Meta additionally revealed a kernel-optimization case examine operating 1,000+ instrument calls over so long as 24 hours.
Is it deployable
Sure. Muse Code ships in beta for macOS and Linux by way of curl -fsSL https://dev.meta.ai/set up.sh | bash. Muse Spark 1.2 is on the market in Muse Code and the Meta Mannequin API, with expanded world entry. The launch submit doesn’t point out downloadable weights, so deal with this as a hosted dependency.
- Firm degree: The API path suits any measurement. The Muse Code path suits groups already operating brokers in sandboxes with evaluate gates.
- Industries: Software program and SaaS, developer tooling, fintech engineering, GPU and inference infrastructure, semiconductors and HPC.
- Functions: Repository-scale refactors and migrations, long-running bug triage, take a look at technology, and GPU kernel optimization.
Async background brokers
Muse Code runs a easy agent loop plus a set of async background brokers. These specialised brokers stay energetic all through every session. They aren’t spawned for particular person duties, which Meta says avoids redundant info gathering. They perform subsequent steps and select when to report again to the primary agent. Meta states this persistence reduces latency and steering on tough, multi-step duties.
Runtime design
Muse Code makes use of a neighborhood occasion log. Each mannequin name, instrument run, approval, and edit is appended to it. Meta calls this single supply of fact replay-exact and restart-safe. After a crash, the agent resumes exactly the place it stopped, letting long-running duties survive failures.
Bundled abilities
Three default abilities ship with the agent. /plan turns a process into an approval-gated plan. /grill stress-tests that plan till it holds up. /purpose works towards profitable completion of the required goal.
What modified in Muse Spark 1.2
Muse Spark 1.2 is a coding-focused replace to Muse Spark 1.1. Meta experiences positive factors in code technology, advanced debugging, codebase understanding, and end-to-end developer workflows. The analysis staff considerably scaled up coaching compute on coding duties and expanded atmosphere variety. The mannequin retains its energy in different areas, together with common brokers.
Three essential coaching particulars:
- Co-training with the harness: Muse Spark 1.2 was co-trained with Muse Code. Coaching included rejection-sampled harness trajectories and recipe optimizations for objectives, compaction, and subagents. The Muse Code toolset was built-in to maximise harness compatibility.
- Lengthy-horizon coaching: Coaching lined whole-repository technology, giant end-to-end tasks, and auto-research. The mannequin makes use of planning, purpose conditioning, and context compaction to maintain progress.
- Self-improvement: Muse Spark 1.1 generated difficult coding environments and instruction-following templates, then graded candidate options towards these necessities. That produced a scalable coaching dataset for 1.2.
Analysis
Meta’s methodology report is unusually particular. Terminal-Bench 2.1 makes use of all 89 duties, go@1 over 5 makes an attempt. DeepSWE v1.1 covers 113 duties throughout 91 repositories and 5 languages. Meta Inner Coding Bench holds 440 duties derived from actual inside pull requests. Runs execute in remoted Daytona cloud sandboxes. Comparisons embody Grok 4.5, Claude Opus 5, GPT-5.6 Terra, Gemini 3.6 Flash, and Kimi K3, every with its personal agent product. Meta notes its harness is probably not tuned for third-party fashions. For reference, Meta’s mannequin web page lists Muse Spark 1.1 at 80.0 on Terminal-Bench 2.1.
Case examine: kernel optimization
Meta examined iterative GPU kernel optimization over 1,000+ instrument calls, operating as much as 24 hours. The mannequin writes, compiles, profiles, and progressively improves kernels towards a offered baseline. Benchmarks lined 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 measurement 1, 64 heads, sequence size 8192, and latent dimension 512. The mannequin constructed a two-kernel Triton pipeline that reuses the shared KV latent as each Okay and V.
Key Takeaways
- Muse Code is a beta terminal coding agent for macOS and Linux, powered by Muse Spark 1.2.
- Persistent async background brokers exchange per-task spawning to chop redundant info gathering.
- An append-only native occasion log makes the runtime replay-exact and restart-safe after crashes.
- Muse Spark 1.2 was co-trained with the harness and skilled on long-horizon, repository-scale work.
- Kernel case examine ran 1,000+ instrument calls over 24 hours on NVIDIA Hopper KDA and MLA kernels.
Try the Technical particulars, Mannequin (Muse Spark 1.2) and Analysis Methodology. Additionally, be at liberty to comply with us on Twitter and don’t overlook to affix our 150k+ML SubReddit and Subscribe to our Publication. Wait! are you on telegram? now you possibly can be a part of us on telegram as nicely.
Have to accomplice with us for selling your GitHub Repo OR Hugging Face Web page OR Product Launch OR Webinar and so on.? Join with us
Asif Razzaq is the CEO of Marktechpost Media Inc.. As a visionary entrepreneur and engineer, Asif is dedicated to harnessing the potential of Synthetic Intelligence for social good. His most up-to-date endeavor is the launch of an Synthetic Intelligence Media Platform, Marktechpost, which stands out for its in-depth protection of machine studying and deep studying information that’s each technically sound and simply comprehensible by a large viewers. The platform boasts of over 2 million month-to-month views, illustrating its reputation amongst audiences.
