{
  "story_id": "8dfc5c29b182cc212a62a3b28981922e",
  "desk": "drm3",
  "revision": 2,
  "published_at": "2026-09-03T16:19:49.000Z",
  "content_hash": "6cb4c303b0b06beb73fe2752d8ab071b161ffa53f07151bc3ae3a4fe636cac74",
  "hash_basis": "sha256 over `headline\\ndek\\nprose`, plus `\\n` + the canonical citations JSON when any source is placed, plus `\\n#blog` for blogs",
  "basis": {
    "headline": "Meta Unveils Muse Spark 1.3 with Enhanced Coding and AI Agent Capabilities",
    "dek": "Meta releases Muse Spark 1.3, a more capable AI model for developers focusing on coding and autonomous tasks.",
    "prose": "Meta is releasing Muse Spark 1.3, a more capable version of its flagship AI model specifically targeted at developers. [^1]\n\nMeta Platforms released Muse Spark 1.3, an AI coding model now available through its Muse Code product and Meta API. [^2] Google and Meta released their latest frontier AI models within hours of each other on Wednesday. [^3] Google shipped Gemini 3.8 Flash alongside a cybersecurity variant, while Meta pushed out Muse Spark 1.3. [^4]\n\nMeta CEO Mark Zuckerberg claimed on X that Muse Spark 1.3 delivers frontier performance and represents the biggest jump yet in coding and agentic work. [^5][^7] Artificial Analysis calculated that Muse Spark 1.3 xhigh costs around $0.55 per Intelligence Index task, the lowest of any model scoring 59 or higher on its index. [^6] Meta Platforms reported that Muse Spark 1.3 scored 98.5 on the MRCR 256K-512K benchmark and 98.1 on MRCR 512K-1M, the highest results in its comparison group for long-context understanding tests. [^8][^12]\n\nMeta states that the Muse Spark 1.3 update pushes deeper into coding and autonomous task completion capabilities. [^9] The company says the Muse Spark 1.3 update improves efficiency in its operations. [^10] Meta Platforms reported that Muse Spark 1.3 scored 75.4 on DeepSWE v1.1, 59.4 on SWEAtlas CodeBase QnA, and 88.8 on Termial-Bench 2.1, leading in coding-related benchmarks, with the Termial-Bench score matching GPT-5.6's result. [^11]",
    "cited": "[{\"statement\":\"Meta is releasing Muse Spark 1.3, a more capable version of its flagship AI model specifically targeted at developers.\",\"source\":\"firstpost.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-02T12:33:00.000Z\",\"publisher_count\":1,\"sources\":[\"firstpost.com\"]},{\"statement\":\"Meta Platforms released Muse Spark 1.3, an AI coding model now available through its Muse Code product and Meta API.\",\"source\":\"Investing.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-02T19:35:35.000Z\",\"publisher_count\":1,\"sources\":[\"Investing.com\"]},{\"statement\":\"Google and Meta released their latest frontier AI models within hours of each other on Wednesday.\",\"source\":\"beincrypto.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-03T05:39:47.000Z\",\"publisher_count\":1,\"sources\":[\"beincrypto.com\"]},{\"statement\":\"Google shipped Gemini 3.8 Flash alongside a cybersecurity variant, while Meta pushed out Muse Spark 1.3.\",\"source\":\"beincrypto.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-03T05:39:47.000Z\",\"publisher_count\":1,\"sources\":[\"beincrypto.com\"]},{\"statement\":\"Meta co-founder and CEO Mark Zuckerberg announced on X that Muse Spark 1.3 is rolling out with frontier performance in coding and agentic work.\",\"source\":\"venturebeat.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-03T16:19:49.000Z\",\"publisher_count\":1,\"sources\":[\"venturebeat.com\"]},{\"statement\":\"Artificial Analysis calculated that Muse Spark 1.3 xhigh costs around $0.55 per Intelligence Index task, the lowest of any model scoring 59 or higher on its index.\",\"source\":\"thenewstack.io\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-03T15:37:07.000Z\",\"publisher_count\":1,\"sources\":[\"thenewstack.io\"]},{\"statement\":\"Meta CEO Mark Zuckerberg claimed on X that Muse Spark 1.3 delivers frontier performance and represents the biggest jump yet in coding and agentic work.\",\"source\":\"thenewstack.io\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-03T15:37:07.000Z\",\"publisher_count\":1,\"sources\":[\"thenewstack.io\"]},{\"statement\":\"Muse Spark 1.3 scored 75.4% on the DeepSWE coding benchmark and 98.1% on the 512K-1M version of the MRCR long-context test according to Meta.\",\"source\":\"thenewstack.io\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-03T15:37:07.000Z\",\"publisher_count\":1,\"sources\":[\"thenewstack.io\"]},{\"statement\":\"Meta states that the Muse Spark 1.3 update pushes deeper into coding and autonomous task completion capabilities.\",\"source\":\"firstpost.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-02T12:33:00.000Z\",\"publisher_count\":1,\"sources\":[\"firstpost.com\"]},{\"statement\":\"The company says the Muse Spark 1.3 update improves efficiency in its operations.\",\"source\":\"firstpost.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-02T12:33:00.000Z\",\"publisher_count\":1,\"sources\":[\"firstpost.com\"]},{\"statement\":\"Meta Platforms reported that Muse Spark 1.3 scored 75.4 on DeepSWE v1.1, 59.4 on SWEAtlas CodeBase QnA, and 88.8 on Termial-Bench 2.1, leading in coding-related benchmarks, with the Termial-Bench score matching GPT-5.6's result.\",\"source\":\"Investing.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-02T19:35:35.000Z\",\"publisher_count\":1,\"sources\":[\"Investing.com\"]},{\"statement\":\"Meta Platforms reported that Muse Spark 1.3 scored 98.5 on the MRCR 256K-512K benchmark and 98.1 on MRCR 512K-1M, the highest results in its comparison group for long-context understanding tests.\",\"source\":\"Investing.com\",\"instrument\":\"News\",\"claim_key\":null,\"published_at\":\"2026-09-02T19:35:35.000Z\",\"publisher_count\":1,\"sources\":[\"Investing.com\"]}]",
    "kind": "news"
  },
  "receipt_verify": "Ed25519 over the dot-joined string `slice_hash.cursor_from.cursor_to.view.view_version.row_count`; public_key and sig are base64url of the raw 32-byte key / 64-byte signature",
  "receipt": null,
  "receipt_note": "this revision predates receipt-keeping (before v0.37.0); the filed row lives in the record",
  "generation_chain": {
    "wire": {
      "stream": "fountain_news",
      "story_id": "7d234a3d6c19a460586695fae94c465a",
      "thread_id": "a39f008a916dd4837818fd4ecf038ae6",
      "thread_label": "Muse Spark 1.3",
      "novelty": "UPDATE",
      "content_hash": "1727508ad328e03e1c08182ea6bb76b282ddbdce739e26eb776fd7be5613ef92",
      "last_published_at": "2026-09-03T16:19:49.000Z",
      "read_receipt": {
        "slice_hash": "f00ea4487a35f7d6cb0b2ecba64018b42f5528b23cf0458226e121b2c25753aa",
        "cursor_from": "eyJ0cyI6IjIwMjYtMDktMDNUMTU6NTg6NDQuMDAwMDAwWiIsImlkIjoiN2JjZTlmODJkMDczOWNkYTMzY2VhYjE5ZGNkOGU5OWYiLCJ2IjoiMSJ9",
        "cursor_to": "eyJ0cyI6IjIwMjYtMDktMDNUMTY6MjI6MjQuMDAwMDAwWiIsImlkIjoiYjQ3NTBjZTc2M2YxODkwMDk4OGYyMWZhYWEzYWUxNjkiLCJ2IjoiMSJ9",
        "view": "v_fountain_news",
        "view_version": "1",
        "row_count": 100,
        "window_days": 3,
        "bytes_scanned": 12166856,
        "credits": 8,
        "price_per_100_rows": 8,
        "sig": "jDh8FniK9MnI4xz1KpeRtnCOCLXOU1SpzEza60xUk_ARXnTeHK9vLPcFsK07K5ZMhoOm6uykIvDilYrWTP-TBQ",
        "public_key": "bMUigy8O0jOnBxQ4Sc-5lwhIZ8LQVAhxMbR7qESVuUE",
        "signer_path": "lakehouse/data-extract/v1",
        "alg": "Ed25519",
        "signed": true
      },
      "compose": 3,
      "recomposed_at": "2026-09-05T20:35:43.353Z"
    },
    "written_at": "2026-09-04T06:32:42.992Z"
  },
  "cited_facts": [
    {
      "n": 1,
      "statement": "Meta is releasing Muse Spark 1.3, a more capable version of its flagship AI model specifically targeted at developers.",
      "source": "firstpost.com",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 2,
      "statement": "Meta Platforms released Muse Spark 1.3, an AI coding model now available through its Muse Code product and Meta API.",
      "source": "Investing.com",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 3,
      "statement": "Google and Meta released their latest frontier AI models within hours of each other on Wednesday.",
      "source": "beincrypto.com",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 4,
      "statement": "Google shipped Gemini 3.8 Flash alongside a cybersecurity variant, while Meta pushed out Muse Spark 1.3.",
      "source": "beincrypto.com",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 5,
      "statement": "Meta co-founder and CEO Mark Zuckerberg announced on X that Muse Spark 1.3 is rolling out with frontier performance in coding and agentic work.",
      "source": "venturebeat.com",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 6,
      "statement": "Artificial Analysis calculated that Muse Spark 1.3 xhigh costs around $0.55 per Intelligence Index task, the lowest of any model scoring 59 or higher on its index.",
      "source": "thenewstack.io",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 7,
      "statement": "Meta CEO Mark Zuckerberg claimed on X that Muse Spark 1.3 delivers frontier performance and represents the biggest jump yet in coding and agentic work.",
      "source": "thenewstack.io",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 8,
      "statement": "Muse Spark 1.3 scored 75.4% on the DeepSWE coding benchmark and 98.1% on the 512K-1M version of the MRCR long-context test according to Meta.",
      "source": "thenewstack.io",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 9,
      "statement": "Meta states that the Muse Spark 1.3 update pushes deeper into coding and autonomous task completion capabilities.",
      "source": "firstpost.com",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 10,
      "statement": "The company says the Muse Spark 1.3 update improves efficiency in its operations.",
      "source": "firstpost.com",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 11,
      "statement": "Meta Platforms reported that Muse Spark 1.3 scored 75.4 on DeepSWE v1.1, 59.4 on SWEAtlas CodeBase QnA, and 88.8 on Termial-Bench 2.1, leading in coding-related benchmarks, with the Termial-Bench score matching GPT-5.6's result.",
      "source": "Investing.com",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    },
    {
      "n": 12,
      "statement": "Meta Platforms reported that Muse Spark 1.3 scored 98.5 on the MRCR 256K-512K benchmark and 98.1 on MRCR 512K-1M, the highest results in its comparison group for long-context understanding tests.",
      "source": "Investing.com",
      "instrument": "News",
      "claim_key": null,
      "url": null,
      "occurred_at": null,
      "published_at": null
    }
  ],
  "note": "A signature proves who filed this and that it has not changed since. It never makes a claim true. The dates on cited_facts (occurred_at, published_at) are the forge's, joined at read time; they are not part of the signed basis above."
}