{"slug": "muse-spark-1-3", "title": "Muse Spark 1.3", "summary": "Muse Spark 1.3, a model trained for long-horizon agentic workflows and competitive coding, was released with a 1M-token context window and benchmark results the company reports ahead of Muse Spark 1.2 on most agentic tasks. On the internal Agentic IF Index, Muse Spark 1.3 (max) scored 98.5 versus 66.3 for Muse Spark 1.2 (xhigh), and it posted 75.4 on DeepSWE v1.1 long-horizon agentic coding against 55.0 for 1.2, while GPT 5.6 Sol (max) scored 73.0 and Opus 5 (max) 74.0 on the same benchmark. The model is priced at $1.25 per million input tokens, $0.15 per million cached input tokens and $4.25 per million output tokens, with a $0.10/$0.002/$0.20 contributor tier whose data is used to improve the company's products.", "body_md": "# Muse Spark 1.3\n\nMuse Spark 1.3 is trained for agentic workflows and optimized for competitive coding performance. Developers can expect higher first-attempt accuracy and reliable tool calling.\n## Meet Muse Spark 1.3\n\nTrained for long-horizon, agentic workflows\n\nCompetitive coding performance\n\nNative multimodal perception\n\nMuse Spark\n\n### Introducing Muse Spark 1.3\n\nIntroducing Muse Spark 1.3, with max reasoning for challenging reasoning and agentic tasks and improved real-world usability.\n## Muse Spark 1.3 benchmarks\n\nCategory\n\nBenchmark\n\nMuse Spark 1.3 (max)\n\nMeta\n\nMuse Spark 1.2 (xhigh)\n\nMeta\n\nGPT 5.6 Sol (max)\n\nOpenAI\n\nOpus 5 (max)\n\nAnthropic\n\nAgent\n\nLong Context\n\nCoding\n\nGDPVal-AA v2Knowledge work\n\nJobBenchProfessional tool use\n\nOSWorld 2.0Agentic computer use\n\nDeepSearchQAAgentic browsing\n\nAgentic IF Index (Internal)Instruction following\n\nAutomationBenchE2E business workflows\n\nMRCR 256K–512KLong context\n\nMRCR 512K–1MLong context\n\nDeepSWE v1.1Long-horizon agentic coding\n\nSWEAtlas CodeBase QnACodebase understanding\n\nTerminal-Bench 2.1Agentic terminal coding\n\n1754\n\n64.9\n\n66.9(partial)32.0(binary)\n\n90.3\n\n57.8\n\n49.6\n\n98.5\n\n98.1\n\n75.4\n\n59.4\n\n88.8\n\n1615\n\n61.6\n\n47.6(partial)17.9(binary)\n\n85.9\n\n46.2\n\n38.2\n\n66.3\n\n55.5\n\n55.0\n\n46.2\n\n82.9\n\n1710\n\n45.4\n\n62.7(partial)27.3(binary)\n\n93.1\n\n60.5\n\n46.7\n\n91.5\n\n73.8\n\n73.0\n\n53.5\n\n88.8\n\n1824\n\n65.7\n\n68.3(partial)31.4(binary)\n\n90.4\n\n59.1\n\n50.3\n\n-\n\n-\n\n74.0\n\n52.7\n\n86.7\n\nGET STARTED\n\n### More ways to build\n\n## Models and pricing\n\nModel\n\nContext window\n\nInput (Mtok)\n\nCached input (Mtok)\n\nOutput (Mtok)\n\nmuse-spark-1.3-contributorUsed to improve our products.\n\n1M\n\n$0.10\n\n$0.002\n\n$0.20\n\nmuse-spark-1.3Not used to improve our products.\n\n1M\n\n$1.25\n\n$0.15\n\n$4.25", "url": "https://wpnews.pro/news/muse-spark-1-3", "canonical_source": "https://dev.meta.ai/models/muse-spark", "published_at": "2026-11-21 03:49:41+00:00", "updated_at": "2026-11-21 04:13:10.309019+00:00", "lang": "en", "topics": ["large-language-models", "ai-agents", "ai-tools", "generative-ai"], "entities": ["Muse Spark 1.3", "Muse Spark 1.2", "GPT 5.6 Sol", "Opus 5", "OpenAI", "Anthropic"], "also_reported_by": [], "takeaway": "Muse Spark 1.3 offers a 1M-token context window at $1.25 per million input and $4.25 per million output tokens, or $0.10/$0.20 on the contributor tier whose data is used for training, and scores 98.5 on the internal Agentic IF Index versus 66.3 for Muse Spark 1.2.", "recommendation": null, "alternates": {"html": "https://wpnews.pro/news/muse-spark-1-3", "markdown": "https://wpnews.pro/news/muse-spark-1-3.md", "text": "https://wpnews.pro/news/muse-spark-1-3.txt", "jsonld": "https://wpnews.pro/news/muse-spark-1-3.jsonld"}}