Meta Releases Muse Spark 1.3 AI Model With Advanced Coding Capabilities

Meta Platforms (NASDAQ:META) has released Muse Spark 1.3, a new version of its artificial intelligence model that the company claims represents its most significant advance yet in coding and agentic work capabilities. CEO Mark Zuckerberg stated the update delivers "frontier performance at low cost," signaling Meta's continued push to compete at the highest levels of the AI industry.

Performance Benchmarks

Meta reported that Muse Spark 1.3 has demonstrated leading performance across several industry benchmarks, particularly in tasks related to software development and long-context understanding. The company highlighted the model's ability to match the performance of competitor models like GPT-5.6 in specific tests.

Key results released by the company include:

• Long-Context Understanding: A score of 98.5 on the MRCR 256K–512K benchmark and 98.1 on MRCR 512K–1M, the highest in its comparison group.
• Coding: Leading scores of 75.4 on DeepSWE v1.1 and 59.4 on SWEAtlas CodeBase QnA. It also achieved 88.8 on Terminal-Bench 2.1, matching the score of GPT-5.6.

Competitive Landscape

Ad

While the model shows strength in coding, benchmark data indicates a mixed performance in agentic capabilities compared to its top rivals. On the GDPVal-AA v2 agent test, Muse Spark 1.3 scored 1754, compared to a score of 1824 for the model known as Opus.

Furthermore, on the DeepSearchQA and Agentic IF Index benchmarks, Muse Spark 1.3 scored 89.4 and 57.8, respectively, trailing GPT-5.6’s scores of 93.0 and 60.5. These figures suggest that while Meta is closing the gap, competitors still hold an edge in certain complex reasoning and automation tasks.

Availability and Market Context

The new model is now accessible to developers through Muse Code and Meta’s API. Zuckerberg also confirmed that Meta plans to release open weights versions in the near future, continuing its strategy of contributing to the open-source AI community.

The announcement comes as competition in the AI sector intensifies, with major technology firms vying for dominance in model performance and developer adoption. Shares of Meta were trading up 2.47% on the day of the release, reflecting positive investor sentiment.