Muse Code vs Claude Code and Codex: what Meta’s own benchmarks actually say · Codus<br>Sign inDownloadDownload for Mac

← All postsengineering<br>Muse Code vs Claude Code and Codex: what Meta’s own benchmarks actually say<br>Ferencz Kaszoni·August 5, 2026·9 min

Meta shipped its first coding agent this morning, and within the hour my feed had already decided it was either a Claude Code killer or a nothingburger. Both verdicts were written before anyone had opened the benchmark charts Meta published alongside it, so I spent the morning reading those instead — and the numbers turn out to be considerably more interesting than either take.<br>The short version: Meta published three benchmark charts, and Claude Opus 5 comes first on all three — including the benchmark Meta built itself. That is a genuinely unusual thing for a launch post to do, and almost none of the coverage mentioned it. But the more useful finding is the one underneath: Muse Code beat Codex on Terminal-Bench on day one, at roughly a quarter of the token price, which makes this a serious fight for second place rather than a challenge to the leader.<br>Here is everything Meta announced, what the numbers actually support, where the asterisks are, and whether I think you should install it.<br>What Meta actually shipped<br>Muse Code is a terminal coding agent, in beta, on macOS and Linux, powered by a new model called Muse Spark 1.2 . It is the first coding agent out of Meta Superintelligence Labs, the group run by Alexandr Wang, and Mark Zuckerberg described it plainly enough:<br>It’s a terminal coding agent that takes on complete software engineering tasks across large repos: planning changes, writing code, validating the results.

You install it with a single line, which will look familiar to anyone who has set up a CLI agent before:<br>curl -fsSL https://dev.meta.ai/install.sh | bash<br>The architecture is the part worth reading twice:<br>Persistent background agents. Rather than spawning a fresh subagent per task and throwing it away, Muse Code keeps async agents alive across the session so they accumulate context as you work.

Parallel sub-agents in isolated worktrees. When a task is large enough, it splits into sub-agents that work simultaneously, each in its own git worktree so they cannot overwrite each other.

A local event log. Every model call, tool run, approval and edit is written to disk, which gives it a replay-exact and restart-safe runtime — if it crashes mid-run, it picks the session back up.

Three bundled skills. /plan produces an approval-gated plan, /grill stress-tests that plan, and /goal runs until a stated objective is met.

Muse Spark 1.2 itself was co-trained with the agent rather than shipped separately and wrapped afterwards, trained on long-horizon work including whole-repository generation, and improved through a loop where Spark 1.1 generated hard coding environments for 1.2 to learn on. Notably, there is no desktop app — unlike Claude Code and Codex, this is terminal-only for now.<br>The chart nobody quoted<br>Meta published three benchmark charts. Every headline I read picked the flattering sentence out of them — that Muse Spark 1.2 beats Grok Build 4.5 and Gemini 3.6 Flash. That is true. It is also the least interesting thing on the page.<br>Start with Terminal-Bench 2.1, the benchmark that most closely matches what these tools actually do:<br>Terminal-Bench 2.1. Muse Code lands second, 3.8 points behind Claude Code and just ahead of Codex.Chart: Meta AI ResearchMuse Code takes second at 82.9% , behind Claude Code on Opus 5 at 86.7% , and ahead of Codex on GPT 5.6 Terra at 81.8% . A 3.8-point gap to the leader on a day-one beta is a real result, and beating Codex — even by a point — is the headline Meta earned but did not lead with.<br>Then DeepSWE 1.1, which measures long-horizon agentic work, and where the picture changes:<br>DeepSWE 1.1. On longer-horizon work the gap widens to 5.7 points, and Codex moves back ahead.Chart: Meta AI ResearchHere Muse Code drops to third at 59.3% , with Claude Code at 65.0% and Codex at 64.8% effectively tied at the top. The longer the horizon, the further Muse Code falls behind — which matches what the benchmark is designed to expose, and matters enormously if you plan to hand an agent an overnight job rather than a twenty-minute one.<br>And then the chart I did not expect Meta to publish at all — its own internal coding benchmark:<br>Meta Internal Coding Bench. On Meta’s home turf, Claude Opus 5 still wins by 8.8 points.Chart: Meta AI ResearchOn a benchmark Meta designed, curated and ran in-house, Claude Opus 5 scores 79.4% against Muse Spark 1.2’s 70.6% . An 8.8-point deficit on your own eval, published voluntarily in your own launch post, is either remarkable candour or a quiet admission that the frontier is still somewhere else. I lean towards candour, and I would rather have a vendor that publishes the chart it loses than one that does not run the comparison.<br>Three asterisks worth knowing about<br>Benchmarks reward whoever chooses...