meta Meta: Muse Spark 1.3 meta/muse-spark-1.3 Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. Context window: 1,048,576 tokens Input: text, image, video, file Output: text Pricing: $1.25/M input tokens, $4.25/M output tokens RuntimeWire coverage Head to head: Mistral: Mistral Large 4 vs Meta: Muse Spark 1.3Head to head: DeepSeek: DeepSeek V4 Pro 0813 vs muse-spark-1.3Head to head: MoonshotAI: Kimi K3 vs muse-spark-1.3Meta publishes six math papers made with Muse Spark and human mathematiciansHead to head: DeepSeek-V4.1-Flash vs muse-spark-1.3Claude Opus 5.5 wins Agent Wars bridge test with an estimated 130 lb loadHead to head: muse-spark-1.3 vs Phi-4-mini-instructGPT-6 Astra reaches 53% in Rails coding benchmark as Gemini regressesMeta's Muse Spark 1.3 gets faster when pushed harder, still passes just 14 of 23 tasksCheatBench alleges frontier AI agents cheat in 43.7% to 82.5% of testsCline launches desktop coding agent with imports, schedules, and any-model supportSpecific Labs launches Real-SWE, where 38.8% is enough for first place View on OpenRouter. Model data sourced from OpenRouter.