Meta's New Muse Spark Model Reportedly Reaches Coding Parity with GPT-5.6
alexandr_wang · x · 2026-08-06
A user on X reported that Meta's newly released Muse Spark 1.2 model has achieved parity with Fable and GPT-5.6 in coding capabilities. In the deepSWE benchmark, the model reportedly outperformed Gemini 3.6 Flash and Grok 4.5.
The poster noted that this is Meta's second major model release in under four weeks, attributing the success to compute power and Mark Zuckerberg's aggressive investments in data centers.
Related event: Meta's New Muse Spark Model Shows Impressive Coding Capabilities(2 posts)→
More from Models
- Meta Launches Muse Code Beta, Its First Coding Agent Powered by Muse Spark 1.2 — alexandr_wang · 2026-08-06
- Muse Spark 1.2 Hits 200 TPS in Tests, Priced at a Fraction of DeepSeek — alexandr_wang · 2026-08-06
- Maple 20B Hits 9,885 tokens/s on Single GH200 with 64 Concurrent Requests — MaziyarPanahi · 2026-08-06
- Gemma 4 Runs Offline on iPhone with Just 500MB of RAM, Outsmarting Siri — cyb3rops · 2026-08-06
- Minimax H3 video generation stuck in 'uncanny valley', dev says — mattshumer_ · 2026-08-06
- DeepSeek V4-Flash tops AI models in affordability, beating Kimi and US rivals — LinkedInNews · 2026-08-06