Meta has released Muse Spark 1.2, an updated version of its coding model, with a new pricing tier that offers 20 cents per million output tokens. This tier requires users to share their data for model improvement, positioning Meta as a price competitor in the market. The model is primarily a coding upgrade to the earlier Muse Spark 1.1, which was released earlier this year. According to Meta, the improvements focus on code generation, debugging, and reasoning over large codebases. The company claims to have increased compute resources for training on programming tasks and expanded the number of training environments. The model was trained on long-running tasks like generating entire repositories or conducting independent research. To manage these extended sessions, it plans steps ahead, works toward a fixed goal, and compresses prior context instead of cutting it off. Some of the training data came from the predecessor model itself. Muse Spark 1.2 shows a clear step up from Spark 1.1 but does not always close the gap to top performers. Image: Meta
Meta cites Terminal-Bench 2.1, DeepSWE v1.1, and 440 tasks from its own codebase as evidence, comparing Spark 1.2 against Grok 4.5, Claude Opus 5, GPT-5.6 Terra, and Gemini 3.6 Flash. Spark 1.2 shows a clear step up from Spark 1.1 but doesn't always close the gap to the top performers. Image: Meta Meta's own methodology acknowledges that the test setup wasn't tuned for competing models and may not reflect their best results. A look at other leaderboards confirms this: Opus 5, for instance, scores about two percentage points higher elsewhere than in Meta's charts. Kimi K3 also shows up in Meta's methodology document. The company says it tested against the open model, but K3 is missing from the published benchmarks. On Terminal-Bench 2.1, K3 trails Opus 5 by only a slim margin and sits well ahead of Spark 1.2. The DeepSWE runs also can't be compared directly to the official leaderboard because each model ran inside its own agent. That's because Meta now ships its own alternative to Claude Code and OpenAI's Codex alongside the new model. The agent picks up exactly where it left off after a crash Muse Code runs in the terminal and installs with a single command, just like Claude Code and OpenAI's Codex. The planning mode will feel familiar too: '/plan' generates a plan that the user has to approve. Less common is the counterpart '/grill,' which stress-tests the plan for weaknesses before execution. There's also '/goal,' which drives the agent toward a fixed objective. The real new feature is how the tool handles sub-agents. Meta's helper agents stay active for an entire session and report back on their own when they have something to share, rather than spinning up for a subtask and shutting down afterward. This should cut down on redundant research. Muse Code also has a fast-resume feature for crashes. The agent logs every model call, every approval, and every change to a local protocol file. Competitors keep full logs too, but Meta claims more precision: after a crash, Muse Code picks up exactly where it stopped instead of re-reading the full context.
Source: thedecoder
