🐦 Twitter Post Details

Viewing enriched Twitter post

@jerryjliu0

Claude Fable 5 thinks document parsing is beneath it It is absolutely crushing on all reasoning-intensive/long horizon benchmarks: SWE-Bench Pro, FrontierCode, GDPval, Runescape, etc. But for document understanding tasks, it is roughly equivalent with Gemini 3 Flash in performance, at roughly 10-15x the token cost. We benchmarked the model on ParseBench and compared it against all other frontier models. It is definitely up there compared to other frontier models, but falls far short of specialized OCR providers. What we found interesting is that Fable 5 is self-aware about this. When we ask the model what tasks it enjoys the last, it actively said that it dislikes tasks "where the request is fully specified and the answer is fully known" - implying part of it being bad is due to laziness and lack of willingness to actually solve the task at hand. For a full list of results across different frontier models, check out ParseBench! https://t.co/PWczfhp0OX

Media 1

šŸ“Š Media Metadata

{
  "media": [
    {
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2064519456966205905/media_0.jpg",
      "media_url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2064519456966205905/media_0.jpg",
      "type": "photo",
      "filename": "media_0.jpg"
    }
  ],
  "processed_at": "2026-06-11T22:53:33.486122",
  "pipeline_version": "2.0"
}

šŸ”§ Raw API Response

{
  "type": "tweet",
  "id": "2064519456966205905",
  "url": "https://x.com/jerryjliu0/status/2064519456966205905",
  "twitterUrl": "https://twitter.com/jerryjliu0/status/2064519456966205905",
  "text": "Claude Fable 5 thinks document parsing is beneath it\n\nIt is absolutely crushing on all reasoning-intensive/long horizon benchmarks: SWE-Bench Pro, FrontierCode,  GDPval, Runescape, etc.\n\nBut for document understanding tasks, it is roughly equivalent with Gemini 3 Flash in performance, at roughly 10-15x the token cost. \n\nWe benchmarked the model on ParseBench and compared it against all other frontier models. It is definitely up there compared to other frontier models, but falls far short of specialized OCR providers.\n\nWhat we found interesting is that Fable 5 is self-aware about this. When we ask the model what tasks it enjoys the last, it actively said that it dislikes tasks \"where the request is fully specified and the answer is fully known\" - implying part of it being bad is due to laziness and lack of willingness to actually solve the task at hand. \n\nFor a full list of results across different frontier models, check out ParseBench! https://t.co/PWczfhp0OX",
  "source": "Twitter for iPhone",
  "retweetCount": 21,
  "replyCount": 24,
  "likeCount": 216,
  "quoteCount": 7,
  "viewCount": 46683,
  "createdAt": "Wed Jun 10 01:26:10 +0000 2026",
  "lang": "en",
  "bookmarkCount": 102,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "2064519456966205905",
  "displayTextRange": [
    0,
    272
  ],
  "inReplyToUserId": null,
  "inReplyToUsername": null,
  "author": {
    "type": "user",
    "userName": "jerryjliu0",
    "url": "https://x.com/jerryjliu0",
    "twitterUrl": "https://twitter.com/jerryjliu0",
    "id": "369777416",
    "name": "Jerry Liu",
    "isVerified": false,
    "isBlueVerified": true,
    "verifiedType": null,
    "profilePicture": "https://pbs.twimg.com/profile_images/1283610285031460864/1Q4zYhtb_normal.jpg",
    "coverPicture": "",
    "description": "",
    "location": "",
    "followers": 76223,
    "following": 1488,
    "status": "",
    "canDm": true,
    "canMediaTag": true,
    "createdAt": "Wed Sep 07 22:54:31 +0000 2011",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 8992,
    "hasCustomTimelines": true,
    "isTranslator": false,
    "mediaCount": 1549,
    "statusesCount": 7124,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [
      "2064002141921112166"
    ],
    "profile_bio": {
      "description": "Parsing the world's hardest PDFs @llama_index. cofounder/CEO\n\nCareers: https://t.co/EUnMNmbCtx\nEnterprise: https://t.co/Ht5jwxSrQB",
      "entities": {
        "description": {
          "urls": [
            {
              "display_url": "llamaindex.ai/careers",
              "expanded_url": "https://www.llamaindex.ai/careers",
              "indices": [
                71,
                94
              ],
              "url": "https://t.co/EUnMNmbCtx"
            },
            {
              "display_url": "llamaindex.ai/contact",
              "expanded_url": "https://www.llamaindex.ai/contact",
              "indices": [
                107,
                130
              ],
              "url": "https://t.co/Ht5jwxSrQB"
            }
          ],
          "user_mentions": [
            {
              "id_str": "",
              "indices": [
                33,
                45
              ],
              "name": "",
              "screen_name": "llama_index"
            }
          ]
        },
        "url": {
          "urls": [
            {
              "display_url": "llamaindex.ai",
              "expanded_url": "https://www.llamaindex.ai/",
              "indices": [
                0,
                23
              ],
              "url": "https://t.co/YiIfjVlzb6"
            }
          ]
        }
      }
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {
    "media": [
      {
        "display_url": "pic.twitter.com/MKkpX2DTRx",
        "expanded_url": "https://twitter.com/jerryjliu0/status/2064519456966205905/photo/1",
        "ext_master_playlist_only": [],
        "ext_media_availability": {
          "status": "Available"
        },
        "ext_playlists": [],
        "features": {
          "large": {
            "faces": []
          },
          "orig": {
            "faces": []
          }
        },
        "id_str": "2064518465986437120",
        "indices": [
          273,
          296
        ],
        "media_key": "3_2064518465986437120",
        "media_results": {
          "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARympJtFWyAACgACHKalggBaYdEAAA==",
          "result": {
            "__typename": "ApiMedia",
            "id": "QXBpTWVkaWE6DAABCgABHKakm0VbIAAKAAIcpqWCAFph0QAA",
            "media_key": "3_2064518465986437120"
          }
        },
        "media_url_https": "https://pbs.twimg.com/media/HKakm0VbIAAwx3G.png",
        "original_info": {
          "focus_rects": [
            {
              "h": 245,
              "w": 438,
              "x": 0,
              "y": 0
            },
            {
              "h": 245,
              "w": 245,
              "x": 0,
              "y": 0
            },
            {
              "h": 245,
              "w": 215,
              "x": 0,
              "y": 0
            },
            {
              "h": 245,
              "w": 123,
              "x": 23,
              "y": 0
            },
            {
              "h": 245,
              "w": 679,
              "x": 0,
              "y": 0
            }
          ],
          "height": 245,
          "width": 679
        },
        "sizes": {
          "large": {
            "h": 245,
            "w": 679
          }
        },
        "type": "photo",
        "url": "https://t.co/MKkpX2DTRx"
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "hashtags": [],
    "symbols": [],
    "urls": [
      {
        "display_url": "parsebench.ai",
        "expanded_url": "https://www.parsebench.ai/",
        "indices": [
          950,
          973
        ],
        "url": "https://t.co/PWczfhp0OX"
      }
    ],
    "user_mentions": []
  },
  "quoted_tweet": {
    "type": "tweet",
    "id": "2064502330800615903",
    "url": "https://x.com/llama_index/status/2064502330800615903",
    "twitterUrl": "https://twitter.com/llama_index/status/2064502330800615903",
    "text": "Day 0 Anthropic Fable 5 in ParseBench: We tested the model's advancements when it comes to document understanding. The model clearly peaks when it comes to adherence to the original text:\n\nšŸ“ƒ Content faithfulness: 90.02% vs 86.19% (Gemini 3 Flash) and 86.81% (GPT-5.5)\nšŸ”¢ Semantic formatting: 72.62% vs 58.35% and 60.12%, a 12+ point lead\n\nThese are two of the most important metrics for SOTA document understanding: does the output preserve what the document actually says, and does it preserve formatting that carries meaning? \n\nBut ... it's not a sweep there continues to be a lot of alpha in unlocking document understanding for frontier models.\n\nFull results below šŸ‘‡",
    "source": "Twitter for iPhone",
    "retweetCount": 7,
    "replyCount": 14,
    "likeCount": 58,
    "quoteCount": 1,
    "viewCount": 37430,
    "createdAt": "Wed Jun 10 00:18:06 +0000 2026",
    "lang": "en",
    "bookmarkCount": 17,
    "isReply": false,
    "inReplyToId": null,
    "conversationId": "2064502330800615903",
    "displayTextRange": [
      0,
      278
    ],
    "inReplyToUserId": null,
    "inReplyToUsername": null,
    "author": {
      "type": "user",
      "userName": "llama_index",
      "url": "https://x.com/llama_index",
      "twitterUrl": "https://twitter.com/llama_index",
      "id": "1604278358296055808",
      "name": "LlamaIndex šŸ¦™",
      "isVerified": false,
      "isBlueVerified": true,
      "verifiedType": "Business",
      "profilePicture": "https://pbs.twimg.com/profile_images/1967920417760251904/0ytfduMQ_normal.png",
      "coverPicture": "https://pbs.twimg.com/profile_banners/1604278358296055808/1776721175",
      "description": "",
      "location": "",
      "followers": 115794,
      "following": 32,
      "status": "",
      "canDm": false,
      "canMediaTag": true,
      "createdAt": "Sun Dec 18 00:52:44 +0000 2022",
      "entities": {
        "description": {
          "urls": []
        },
        "url": {}
      },
      "fastFollowersCount": 0,
      "favouritesCount": 1569,
      "hasCustomTimelines": true,
      "isTranslator": false,
      "mediaCount": 1899,
      "statusesCount": 3903,
      "withheldInCountries": [],
      "affiliatesHighlightedLabel": {},
      "possiblySensitive": false,
      "pinnedTweetIds": [],
      "profile_bio": {
        "description": "The world's best AI Document OCR\n\nLlamaParse: https://t.co/yQGTiRSfFL\nDocs: https://t.co/us6GCS14vD",
        "entities": {
          "description": {
            "urls": [
              {
                "display_url": "cloud.llamaindex.ai",
                "expanded_url": "https://cloud.llamaindex.ai/",
                "indices": [
                  46,
                  69
                ],
                "url": "https://t.co/yQGTiRSfFL"
              },
              {
                "display_url": "developers.llamaindex.ai/python/cloud/",
                "expanded_url": "https://developers.llamaindex.ai/python/cloud/",
                "indices": [
                  76,
                  99
                ],
                "url": "https://t.co/us6GCS14vD"
              }
            ]
          },
          "url": {
            "urls": [
              {
                "display_url": "llamaindex.ai",
                "expanded_url": "https://www.llamaindex.ai/",
                "indices": [
                  0,
                  23
                ],
                "url": "https://t.co/epzefqPT9Z"
              }
            ]
          }
        }
      },
      "isAutomated": false,
      "automatedBy": null
    },
    "extendedEntities": {
      "media": [
        {
          "display_url": "pic.twitter.com/HtyOI2jvoR",
          "expanded_url": "https://twitter.com/llama_index/status/2064502330800615903/photo/1",
          "ext_master_playlist_only": [],
          "ext_media_availability": {
            "status": "Available"
          },
          "ext_playlists": [],
          "features": {
            "large": {
              "faces": []
            },
            "orig": {
              "faces": []
            }
          },
          "id_str": "2064502328116342784",
          "indices": [
            279,
            302
          ],
          "media_key": "3_2064502328116342784",
          "media_results": {
            "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARymle3hV3AACgACHKaV7oFWMd8AAA==",
            "result": {
              "__typename": "ApiMedia",
              "id": "QXBpTWVkaWE6DAABCgABHKaV7eFXcAAKAAIcppXugVYx3wAA",
              "media_key": "3_2064502328116342784"
            }
          },
          "media_url_https": "https://pbs.twimg.com/media/HKaV7eFXcAAKG2S.png",
          "original_info": {
            "focus_rects": [
              {
                "h": 794,
                "w": 1418,
                "x": 0,
                "y": 0
              },
              {
                "h": 794,
                "w": 794,
                "x": 0,
                "y": 0
              },
              {
                "h": 794,
                "w": 696,
                "x": 0,
                "y": 0
              },
              {
                "h": 794,
                "w": 397,
                "x": 77,
                "y": 0
              },
              {
                "h": 794,
                "w": 2200,
                "x": 0,
                "y": 0
              }
            ],
            "height": 794,
            "width": 2200
          },
          "sizes": {
            "large": {
              "h": 739,
              "w": 2048
            }
          },
          "type": "photo",
          "url": "https://t.co/HtyOI2jvoR"
        }
      ]
    },
    "card": null,
    "place": {},
    "entities": {
      "hashtags": [],
      "symbols": [],
      "urls": [],
      "user_mentions": []
    },
    "quoted_tweet": null,
    "retweeted_tweet": null,
    "isLimitedReply": false,
    "communityInfo": null,
    "article": null
  },
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "communityInfo": null,
  "article": null
}