Meta enters the AI coding wars with Muse Spark 1.2 and Muse Code with persistent async background agents - Venturebeat
First reporting: names both products and highlights Muse Code's persistent async background-agent design.
5 items · 3 sources · 3 days
Operational story trace
Follow in this browser to see new updates on your Live feed.
Latest change
Meta open-sourced Muse Glimmer on Aug 14 — a 30B-parameter, Apache 2.0-licensed model built for on-device agentic execution — the first concrete Muse-family spec disclosure since the still-unpriced Aug 5 Muse Code launch.
Meta had no dedicated AI coding-agent product until Aug 5, 2026, when it introduced Muse Code alongside an updated Muse Spark 1.2, framed explicitly as competition for Claude Code and Codex. Coverage repeated the same launch framing through Aug 9, with no pricing, availability, or benchmark details disclosed for either product.
State over time
First reporting: names both products and highlights Muse Code's persistent async background-agent design.
Independent outlet corroborates the same-day Claude Code/Codex comparison framing.
First head-to-head framing against Claude Code and Codex.
Coverage continues four days later with no new product details disclosed.
Meta's first concrete Muse-family release: an Apache 2.0, 30B-parameter model built for on-device agent execution, in contrast to the still-undisclosed Muse Code.
First reporting: names both products and highlights Muse Code's persistent async background-agent design.
First head-to-head framing against Claude Code and Codex.
Independent outlet corroborates the same-day Claude Code/Codex comparison framing.
Coverage continues four days later with no new product details disclosed.
Meta's first concrete Muse-family release: an Apache 2.0, 30B-parameter model built for on-device agent execution, in contrast to the still-undisclosed Muse Code.
What to watch — open questions
Storylines are threaded mechanically from the feed: stories that share a distinctive anchor across multiple days and sources. Each item links to its original source. The evidence trace, current state, and open questions are written by the editor routine and refreshed whenever a new beat lands.