Muse Spark 1.3

Changes from previous version

Follows Muse Spark 1.1 after the 1.2 coding-focused update. Meta engineers report about 20% fewer tool calls and 25% fewer tokens than 1.2, with fewer unneeded turns and cleaner coding style. Long-context retrieval jumps from 66.3% to 98.5% on MRCR 256K-512K and from 55.5% to 98.1% at 512K-1M. Stronger collaboration: asks clarifying questions, confirms before irreversible actions, and maps new prompts to the right task in messy threads.

Release Summary

Meta Superintelligence Labs' latest Muse Spark model for long-horizon agentic work, multitasking in a single thread, and coding. Available today in Muse Code and the Meta Model API with max reasoning. Meta reports 75.4% on DeepSWE v1.1, 59.4% on SWEAtlas CodeBase QnA, 88.8% on Terminal-Bench 2.1 (tied with GPT Sol 5.6), and 98.5% / 98.1% on MRCR at 256K-512K and 512K-1M. Standard API pricing is $1.25/M input and $4.25/M output ($0.15/M cache reads), matching Cursor's listed rates. Measured price per task is $1.60 at max reasoning.

Timeline

Other Muse Spark versions