RubyGems - llm.rb - Versions diffs - 12.0.0 → 12.2.0 - Mend

llm.rb 12.0.0 → 12.2.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Files changed (45) hide show

checksums.yaml +4 -4
data/CHANGELOG.md +193 -1
data/LICENSE +5 -4
data/README.md +56 -18
data/data/anthropic.json +85 -372
data/data/bedrock.json +498 -3
data/data/deepinfra.json +567 -47
data/data/deepseek.json +4 -0
data/data/google.json +54 -0
data/data/mistral.json +968 -0
data/data/openai.json +51 -0
data/data/xai.json +67 -0
data/data/zai.json +14 -0
data/lib/llm/agent.rb +41 -1
data/lib/llm/context.rb +9 -1
data/lib/llm/function.rb +6 -0
data/lib/llm/json_adapter.rb +29 -3
data/lib/llm/provider.rb +3 -3
data/lib/llm/providers/anthropic.rb +2 -2
data/lib/llm/providers/google.rb +3 -1
data/lib/llm/providers/llamacpp.rb +5 -5
data/lib/llm/providers/mistral/request_adapter/completion.rb +122 -0
data/lib/llm/providers/mistral/request_adapter.rb +20 -0
data/lib/llm/providers/mistral.rb +96 -0
data/lib/llm/repl/input.rb +64 -0
data/lib/llm/repl/status.rb +30 -0
data/lib/llm/repl/stream.rb +46 -0
data/lib/llm/repl/transcript.rb +61 -0
data/lib/llm/repl/window.rb +107 -0
data/lib/llm/repl.rb +78 -0
data/lib/llm/tools/chdir.rb +23 -0
data/lib/llm/tools/git.rb +41 -0
data/lib/llm/tools/mkdir.rb +32 -0
data/lib/llm/tools/pwd.rb +20 -0
data/lib/llm/tools/read_file.rb +40 -0
data/lib/llm/tools/rg.rb +46 -0
data/lib/llm/tools/shell.rb +48 -0
data/lib/llm/tools/swap_text.rb +25 -0
data/lib/llm/tools/write_file.rb +24 -0
data/lib/llm/tools.rb +5 -0
data/lib/llm/version.rb +1 -1
data/lib/llm.rb +9 -0
data/llm.gemspec +1 -1
data/resources/deepdive.md +32 -0
metadata +22 -2

data/data/bedrock.json CHANGED Viewed

@@ -13,6 +13,7 @@
     "global.anthropic.claude-haiku-4-5-20251001-v1:0": {
       "id": "global.anthropic.claude-haiku-4-5-20251001-v1:0",
       "name": "Claude Haiku 4.5 (Global)",
+      "description": "Fast Claude model for responsive assistance, classification, and lightweight agents",
       "family": "claude-haiku",
       "attachment": true,
       "reasoning": true,
@@ -53,6 +54,7 @@
     "global.anthropic.claude-sonnet-4-5-20250929-v1:0": {
       "id": "global.anthropic.claude-sonnet-4-5-20250929-v1:0",
       "name": "Claude Sonnet 4.5 (Global)",
+      "description": "Balanced Claude model for coding, analysis, agent workflows, and cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -90,9 +92,51 @@
         "cache_write": 3.75
       }
     },
+    "jp.anthropic.claude-haiku-4-5-20251001-v1:0": {
+      "id": "jp.anthropic.claude-haiku-4-5-20251001-v1:0",
+      "name": "Claude Haiku 4.5 (JP)",
+      "description": "Fast Claude model for responsive assistance, classification, and lightweight agents",
+      "family": "claude-haiku",
+      "attachment": true,
+      "reasoning": true,
+      "reasoning_options": [
+        {
+          "type": "budget_tokens",
+          "min": 1024
+        }
+      ],
+      "tool_call": true,
+      "structured_output": true,
+      "temperature": true,
+      "knowledge": "2025-02-28",
+      "release_date": "2025-10-15",
+      "last_updated": "2025-10-15",
+      "modalities": {
+        "input": [
+          "text",
+          "image",
+          "pdf"
+        ],
+        "output": [
+          "text"
+        ]
+      },
+      "open_weights": false,
+      "limit": {
+        "context": 200000,
+        "output": 64000
+      },
+      "cost": {
+        "input": 1,
+        "output": 5,
+        "cache_read": 0.1,
+        "cache_write": 1.25
+      }
+    },
     "us.meta.llama4-scout-17b-instruct-v1:0": {
       "id": "us.meta.llama4-scout-17b-instruct-v1:0",
       "name": "Llama 4 Scout 17B Instruct (US)",
+      "description": "Open Llama with long-context vision for efficient multimodal agents",
       "family": "llama",
       "attachment": true,
       "reasoning": false,
@@ -123,6 +167,7 @@
     "minimax.minimax-m2": {
       "id": "minimax.minimax-m2",
       "name": "MiniMax M2",
+      "description": "MiniMax model for chat, coding, office work, and agentic tasks",
       "family": "minimax",
       "attachment": false,
       "reasoning": true,
@@ -153,6 +198,7 @@
     "anthropic.claude-opus-4-7": {
       "id": "anthropic.claude-opus-4-7",
       "name": "Claude Opus 4.7",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -198,6 +244,7 @@
     "eu.anthropic.claude-sonnet-4-6": {
       "id": "eu.anthropic.claude-sonnet-4-6",
       "name": "Claude Sonnet 4.6 (EU)",
+      "description": "Claude workhorse for coding agents, careful analysis, and production cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -247,6 +294,7 @@
     "mistral.voxtral-small-24b-2507": {
       "id": "mistral.voxtral-small-24b-2507",
       "name": "Voxtral Small 24B 2507",
+      "description": "Efficient Mistral model for fast chat, extraction, and production assistants",
       "family": "mistral",
       "attachment": true,
       "reasoning": false,
@@ -277,6 +325,7 @@
     "mistral.ministral-3-3b-instruct": {
       "id": "mistral.ministral-3-3b-instruct",
       "name": "Ministral 3 3B",
+      "description": "Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads",
       "family": "ministral",
       "attachment": false,
       "reasoning": false,
@@ -307,6 +356,7 @@
     "openai.gpt-oss-20b": {
       "id": "openai.gpt-oss-20b",
       "name": "gpt-oss-20b",
+      "description": "Open-weight GPT model for self-hosted reasoning and instruction-following workloads",
       "family": "gpt-oss",
       "attachment": false,
       "reasoning": true,
@@ -351,6 +401,7 @@
     "anthropic.claude-opus-4-6-v1": {
       "id": "anthropic.claude-opus-4-6-v1",
       "name": "Claude Opus 4.6",
+      "description": "High-end Claude for difficult coding, planning, and slower expert reasoning",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -400,6 +451,7 @@
     "openai.gpt-oss-safeguard-20b": {
       "id": "openai.gpt-oss-safeguard-20b",
       "name": "GPT OSS Safeguard 20B",
+      "description": "Safety model for policy screening, moderation, and risk-aware routing workflows",
       "family": "gpt-oss",
       "attachment": false,
       "reasoning": false,
@@ -429,6 +481,7 @@
     "anthropic.claude-opus-4-5-20251101-v1:0": {
       "id": "anthropic.claude-opus-4-5-20251101-v1:0",
       "name": "Claude Opus 4.5",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -477,6 +530,7 @@
     "global.anthropic.claude-fable-5": {
       "id": "global.anthropic.claude-fable-5",
       "name": "Claude Fable 5 (Global)",
+      "description": "Claude model for creative writing, analysis, and controlled agent workflows",
       "family": "claude-fable",
       "attachment": true,
       "reasoning": true,
@@ -500,7 +554,8 @@
       "modalities": {
         "input": [
           "text",
-          "image"
+          "image",
+          "pdf"
         ],
         "output": [
           "text"
@@ -521,6 +576,7 @@
     "openai.gpt-oss-120b-1:0": {
       "id": "openai.gpt-oss-120b-1:0",
       "name": "gpt-oss-120b",
+      "description": "Open-weight GPT model for self-hosted reasoning and instruction-following workloads",
       "family": "gpt-oss",
       "attachment": false,
       "reasoning": true,
@@ -560,6 +616,7 @@
     "anthropic.claude-sonnet-4-5-20250929-v1:0": {
       "id": "anthropic.claude-sonnet-4-5-20250929-v1:0",
       "name": "Claude Sonnet 4.5",
+      "description": "Balanced Claude model for coding, analysis, agent workflows, and cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -600,6 +657,7 @@
     "amazon.nova-pro-v1:0": {
       "id": "amazon.nova-pro-v1:0",
       "name": "Nova Pro",
+      "description": "Flagship model for demanding analysis, coding, and production agent workflows",
       "family": "nova-pro",
       "attachment": true,
       "reasoning": false,
@@ -632,6 +690,7 @@
     "qwen.qwen3-coder-next": {
       "id": "qwen.qwen3-coder-next",
       "name": "Qwen3 Coder Next",
+      "description": "Qwen coding model for software agents, repository edits, and code reasoning",
       "family": "qwen",
       "attachment": false,
       "reasoning": true,
@@ -662,6 +721,7 @@
     "us.anthropic.claude-opus-4-7": {
       "id": "us.anthropic.claude-opus-4-7",
       "name": "Claude Opus 4.7 (US)",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -707,6 +767,7 @@
     "nvidia.nemotron-nano-9b-v2": {
       "id": "nvidia.nemotron-nano-9b-v2",
       "name": "NVIDIA Nemotron Nano 9B v2",
+      "description": "Compact Nemotron model for efficient reasoning and deployable AI agents",
       "family": "nemotron",
       "attachment": false,
       "reasoning": false,
@@ -736,6 +797,7 @@
     "qwen.qwen3-32b-v1:0": {
       "id": "qwen.qwen3-32b-v1:0",
       "name": "Qwen3 32B (dense)",
+      "description": "Qwen instruction model for multilingual chat, reasoning, and tool use",
       "family": "qwen",
       "attachment": false,
       "reasoning": true,
@@ -767,6 +829,7 @@
     "jp.anthropic.claude-sonnet-4-6": {
       "id": "jp.anthropic.claude-sonnet-4-6",
       "name": "Claude Sonnet 4.6 (JP)",
+      "description": "Claude workhorse for coding agents, careful analysis, and production cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -816,6 +879,7 @@
     "deepseek.r1-v1:0": {
       "id": "deepseek.r1-v1:0",
       "name": "DeepSeek-R1",
+      "description": "DeepSeek reasoning model for multi-step analysis, math, coding, and tools",
       "family": "deepseek-thinking",
       "attachment": false,
       "reasoning": true,
@@ -846,6 +910,7 @@
     "mistral.mistral-large-3-675b-instruct": {
       "id": "mistral.mistral-large-3-675b-instruct",
       "name": "Mistral Large 3",
+      "description": "Flagship Mistral model for advanced reasoning, coding, and multilingual work",
       "family": "mistral",
       "attachment": false,
       "reasoning": false,
@@ -876,6 +941,7 @@
     "google.gemma-3-27b-it": {
       "id": "google.gemma-3-27b-it",
       "name": "Google Gemma 3 27B Instruct",
+      "description": "Open Gemma instruction model for efficient chat and self-hosted deployments",
       "family": "gemma",
       "attachment": true,
       "reasoning": false,
@@ -907,6 +973,7 @@
     "anthropic.claude-sonnet-4-6": {
       "id": "anthropic.claude-sonnet-4-6",
       "name": "Claude Sonnet 4.6",
+      "description": "Claude workhorse for coding agents, careful analysis, and production cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -956,6 +1023,7 @@
     "amazon.nova-2-lite-v1:0": {
       "id": "amazon.nova-2-lite-v1:0",
       "name": "Nova 2 Lite",
+      "description": "Multimodal reasoning model for visual analysis, planning, and tool use",
       "family": "nova",
       "attachment": false,
       "reasoning": true,
@@ -999,6 +1067,7 @@
     "openai.gpt-oss-safeguard-120b": {
       "id": "openai.gpt-oss-safeguard-120b",
       "name": "GPT OSS Safeguard 120B",
+      "description": "Safety model for policy screening, moderation, and risk-aware routing workflows",
       "family": "gpt-oss",
       "attachment": false,
       "reasoning": false,
@@ -1028,6 +1097,7 @@
     "mistral.ministral-3-8b-instruct": {
       "id": "mistral.ministral-3-8b-instruct",
       "name": "Ministral 3 8B",
+      "description": "Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads",
       "family": "ministral",
       "attachment": false,
       "reasoning": false,
@@ -1057,6 +1127,7 @@
     "eu.anthropic.claude-opus-4-6-v1": {
       "id": "eu.anthropic.claude-opus-4-6-v1",
       "name": "Claude Opus 4.6 (EU)",
+      "description": "High-end Claude for difficult coding, planning, and slower expert reasoning",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -1106,6 +1177,7 @@
     "au.anthropic.claude-opus-4-6-v1": {
       "id": "au.anthropic.claude-opus-4-6-v1",
       "name": "AU Anthropic Claude Opus 4.6",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -1152,9 +1224,60 @@
         "cache_write": 20.625
       }
     },
+    "jp.anthropic.claude-sonnet-5": {
+      "id": "jp.anthropic.claude-sonnet-5",
+      "name": "Claude Sonnet 5 (JP)",
+      "description": "Everyday Claude agent model for coding, planning, browsing, and general work",
+      "family": "claude-sonnet",
+      "attachment": true,
+      "reasoning": true,
+      "reasoning_options": [
+        {
+          "type": "toggle"
+        },
+        {
+          "type": "effort",
+          "values": [
+            "low",
+            "medium",
+            "high",
+            "xhigh",
+            "max"
+          ]
+        }
+      ],
+      "tool_call": true,
+      "structured_output": true,
+      "temperature": false,
+      "knowledge": "2026-01-31",
+      "release_date": "2026-06-30",
+      "last_updated": "2026-06-30",
+      "modalities": {
+        "input": [
+          "text",
+          "image",
+          "pdf"
+        ],
+        "output": [
+          "text"
+        ]
+      },
+      "open_weights": false,
+      "limit": {
+        "context": 1000000,
+        "output": 128000
+      },
+      "cost": {
+        "input": 2,
+        "output": 10,
+        "cache_read": 0.2,
+        "cache_write": 2.5
+      }
+    },
     "openai.gpt-oss-120b": {
       "id": "openai.gpt-oss-120b",
       "name": "gpt-oss-120b",
+      "description": "Open-weight GPT model for self-hosted reasoning and instruction-following workloads",
       "family": "gpt-oss",
       "attachment": false,
       "reasoning": true,
@@ -1196,9 +1319,60 @@
         "output": 0.6
       }
     },
+    "global.anthropic.claude-sonnet-5": {
+      "id": "global.anthropic.claude-sonnet-5",
+      "name": "Claude Sonnet 5 (Global)",
+      "description": "Everyday Claude agent model for coding, planning, browsing, and general work",
+      "family": "claude-sonnet",
+      "attachment": true,
+      "reasoning": true,
+      "reasoning_options": [
+        {
+          "type": "toggle"
+        },
+        {
+          "type": "effort",
+          "values": [
+            "low",
+            "medium",
+            "high",
+            "xhigh",
+            "max"
+          ]
+        }
+      ],
+      "tool_call": true,
+      "structured_output": true,
+      "temperature": false,
+      "knowledge": "2026-01-31",
+      "release_date": "2026-06-30",
+      "last_updated": "2026-06-30",
+      "modalities": {
+        "input": [
+          "text",
+          "image",
+          "pdf"
+        ],
+        "output": [
+          "text"
+        ]
+      },
+      "open_weights": false,
+      "limit": {
+        "context": 1000000,
+        "output": 128000
+      },
+      "cost": {
+        "input": 2,
+        "output": 10,
+        "cache_read": 0.2,
+        "cache_write": 2.5
+      }
+    },
     "global.anthropic.claude-opus-4-8": {
       "id": "global.anthropic.claude-opus-4-8",
       "name": "Claude Opus 4.8 (Global)",
+      "description": "Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -1216,6 +1390,7 @@
       ],
       "tool_call": true,
       "temperature": false,
+      "knowledge": "2026-01",
       "release_date": "2026-05-28",
       "last_updated": "2026-05-28",
       "modalities": {
@@ -1243,6 +1418,7 @@
     "us.meta.llama4-maverick-17b-instruct-v1:0": {
       "id": "us.meta.llama4-maverick-17b-instruct-v1:0",
       "name": "Llama 4 Maverick 17B Instruct (US)",
+      "description": "Open multimodal Llama for strong reasoning with efficient everyday serving",
       "family": "llama",
       "attachment": true,
       "reasoning": false,
@@ -1273,6 +1449,7 @@
     "openai.gpt-5.4": {
       "id": "openai.gpt-5.4",
       "name": "GPT-5.4",
+      "description": "Agent-ready GPT for coding and computer-use workflows at a lower cost",
       "family": "gpt",
       "attachment": true,
       "reasoning": true,
@@ -1320,9 +1497,60 @@
         "cache_read": 0.275
       }
     },
+    "anthropic.claude-sonnet-5": {
+      "id": "anthropic.claude-sonnet-5",
+      "name": "Claude Sonnet 5",
+      "description": "Everyday Claude agent model for coding, planning, browsing, and general work",
+      "family": "claude-sonnet",
+      "attachment": true,
+      "reasoning": true,
+      "reasoning_options": [
+        {
+          "type": "toggle"
+        },
+        {
+          "type": "effort",
+          "values": [
+            "low",
+            "medium",
+            "high",
+            "xhigh",
+            "max"
+          ]
+        }
+      ],
+      "tool_call": true,
+      "structured_output": true,
+      "temperature": false,
+      "knowledge": "2026-01-31",
+      "release_date": "2026-06-30",
+      "last_updated": "2026-06-30",
+      "modalities": {
+        "input": [
+          "text",
+          "image",
+          "pdf"
+        ],
+        "output": [
+          "text"
+        ]
+      },
+      "open_weights": false,
+      "limit": {
+        "context": 1000000,
+        "output": 128000
+      },
+      "cost": {
+        "input": 2,
+        "output": 10,
+        "cache_read": 0.2,
+        "cache_write": 2.5
+      }
+    },
     "mistral.devstral-2-123b": {
       "id": "mistral.devstral-2-123b",
       "name": "Devstral 2 123B",
+      "description": "Mistral coding agent model for repository tasks and software engineering workflows",
       "family": "devstral",
       "attachment": false,
       "reasoning": false,
@@ -1352,6 +1580,7 @@
     "zai.glm-4.7": {
       "id": "zai.glm-4.7",
       "name": "GLM-4.7",
+      "description": "Flagship GLM model for hybrid reasoning, coding, and agentic engineering",
       "family": "glm",
       "attachment": false,
       "reasoning": true,
@@ -1383,9 +1612,56 @@
         "output": 2.2
       }
     },
+    "anthropic.claude-fable-5": {
+      "id": "anthropic.claude-fable-5",
+      "name": "Claude Fable 5",
+      "description": "Claude model for creative writing, analysis, and controlled agent workflows",
+      "family": "claude-fable",
+      "attachment": true,
+      "reasoning": true,
+      "reasoning_options": [
+        {
+          "type": "effort",
+          "values": [
+            "low",
+            "medium",
+            "high",
+            "xhigh",
+            "max"
+          ]
+        }
+      ],
+      "tool_call": true,
+      "temperature": false,
+      "knowledge": "2026-01-31",
+      "release_date": "2026-06-09",
+      "last_updated": "2026-06-09",
+      "modalities": {
+        "input": [
+          "text",
+          "image",
+          "pdf"
+        ],
+        "output": [
+          "text"
+        ]
+      },
+      "open_weights": false,
+      "limit": {
+        "context": 1000000,
+        "output": 128000
+      },
+      "cost": {
+        "input": 10,
+        "output": 50,
+        "cache_read": 1,
+        "cache_write": 12.5
+      }
+    },
     "writer.palmyra-x4-v1:0": {
       "id": "writer.palmyra-x4-v1:0",
       "name": "Palmyra X4",
+      "description": "Reasoning model for deliberate analysis, multi-step problem solving, and tool use",
       "family": "palmyra",
       "attachment": false,
       "reasoning": true,
@@ -1415,6 +1691,7 @@
     "mistral.magistral-small-2509": {
       "id": "mistral.magistral-small-2509",
       "name": "Magistral Small 1.2",
+      "description": "Mistral reasoning model for transparent analysis, math, and complex decisions",
       "family": "magistral",
       "attachment": false,
       "reasoning": true,
@@ -1446,6 +1723,7 @@
     "qwen.qwen3-coder-480b-a35b-v1:0": {
       "id": "qwen.qwen3-coder-480b-a35b-v1:0",
       "name": "Qwen3 Coder 480B A35B Instruct",
+      "description": "Qwen coding model for software agents, repository edits, and code reasoning",
       "family": "qwen",
       "attachment": false,
       "reasoning": false,
@@ -1476,6 +1754,7 @@
     "amazon.nova-micro-v1:0": {
       "id": "amazon.nova-micro-v1:0",
       "name": "Nova Micro",
+      "description": "Efficient model for low-latency assistance, extraction, and routine automation",
       "family": "nova-micro",
       "attachment": false,
       "reasoning": false,
@@ -1506,6 +1785,7 @@
     "mistral.pixtral-large-2502-v1:0": {
       "id": "mistral.pixtral-large-2502-v1:0",
       "name": "Pixtral Large (25.02)",
+      "description": "Mistral vision-language model for image understanding and multimodal chat",
       "family": "mistral",
       "attachment": false,
       "reasoning": false,
@@ -1535,6 +1815,7 @@
     "us.anthropic.claude-opus-4-6-v1": {
       "id": "us.anthropic.claude-opus-4-6-v1",
       "name": "Claude Opus 4.6 (US)",
+      "description": "High-end Claude for difficult coding, planning, and slower expert reasoning",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -1584,6 +1865,7 @@
     "jp.anthropic.claude-opus-4-7": {
       "id": "jp.anthropic.claude-opus-4-7",
       "name": "Claude Opus 4.7 (JP)",
+      "description": "Stronger Opus tier for advanced software work and high-stakes reasoning",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -1629,6 +1911,7 @@
     "au.anthropic.claude-sonnet-4-5-20250929-v1:0": {
       "id": "au.anthropic.claude-sonnet-4-5-20250929-v1:0",
       "name": "Claude Sonnet 4.5 (AU)",
+      "description": "Balanced Claude model for coding, analysis, agent workflows, and cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -1669,6 +1952,7 @@
     "deepseek.v3-v1:0": {
       "id": "deepseek.v3-v1:0",
       "name": "DeepSeek-V3.1",
+      "description": "DeepSeek chat model for instruction following, coding, and analysis",
       "family": "deepseek",
       "attachment": false,
       "reasoning": true,
@@ -1700,6 +1984,7 @@
     "anthropic.claude-opus-4-1-20250805-v1:0": {
       "id": "anthropic.claude-opus-4-1-20250805-v1:0",
       "name": "Claude Opus 4.1",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -1739,6 +2024,7 @@
     "jp.anthropic.claude-sonnet-4-5-20250929-v1:0": {
       "id": "jp.anthropic.claude-sonnet-4-5-20250929-v1:0",
       "name": "Claude Sonnet 4.5 (JP)",
+      "description": "Balanced Claude model for coding, analysis, agent workflows, and cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -1779,6 +2065,7 @@
     "google.gemma-3-4b-it": {
       "id": "google.gemma-3-4b-it",
       "name": "Gemma 3 4B IT",
+      "description": "Open Gemma instruction model for efficient chat and self-hosted deployments",
       "family": "gemma",
       "attachment": false,
       "reasoning": false,
@@ -1808,6 +2095,7 @@
     "eu.anthropic.claude-sonnet-4-5-20250929-v1:0": {
       "id": "eu.anthropic.claude-sonnet-4-5-20250929-v1:0",
       "name": "Claude Sonnet 4.5 (EU)",
+      "description": "Balanced Claude model for coding, analysis, agent workflows, and cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -1848,6 +2136,7 @@
     "qwen.qwen3-vl-235b-a22b": {
       "id": "qwen.qwen3-vl-235b-a22b",
       "name": "Qwen/Qwen3-VL-235B-A22B-Instruct",
+      "description": "Qwen vision-language model for visual reasoning, documents, and agent tasks",
       "family": "qwen",
       "attachment": true,
       "reasoning": false,
@@ -1878,6 +2167,7 @@
     "writer.palmyra-x5-v1:0": {
       "id": "writer.palmyra-x5-v1:0",
       "name": "Palmyra X5",
+      "description": "Reasoning model for deliberate analysis, multi-step problem solving, and tool use",
       "family": "palmyra",
       "attachment": false,
       "reasoning": true,
@@ -1907,6 +2197,7 @@
     "us.anthropic.claude-sonnet-4-6": {
       "id": "us.anthropic.claude-sonnet-4-6",
       "name": "Claude Sonnet 4.6 (US)",
+      "description": "Claude workhorse for coding agents, careful analysis, and production cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -1956,6 +2247,7 @@
     "au.anthropic.claude-haiku-4-5-20251001-v1:0": {
       "id": "au.anthropic.claude-haiku-4-5-20251001-v1:0",
       "name": "Claude Haiku 4.5 (AU)",
+      "description": "Fast Claude model for responsive assistance, classification, and lightweight agents",
       "family": "claude-haiku",
       "attachment": true,
       "reasoning": true,
@@ -1996,6 +2288,7 @@
     "meta.llama3-3-70b-instruct-v1:0": {
       "id": "meta.llama3-3-70b-instruct-v1:0",
       "name": "Llama 3.3 70B Instruct",
+      "description": "Open Llama instruction model for multilingual chat, reasoning, and coding",
       "family": "llama",
       "attachment": false,
       "reasoning": false,
@@ -2025,6 +2318,7 @@
     "zai.glm-5": {
       "id": "zai.glm-5",
       "name": "GLM-5",
+      "description": "Flagship GLM model for hybrid reasoning, coding, and agentic engineering",
       "family": "glm",
       "attachment": false,
       "reasoning": true,
@@ -2058,6 +2352,7 @@
     "us.anthropic.claude-opus-4-8": {
       "id": "us.anthropic.claude-opus-4-8",
       "name": "Claude Opus 4.8 (US)",
+      "description": "Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -2075,6 +2370,7 @@
       ],
       "tool_call": true,
       "temperature": false,
+      "knowledge": "2026-01",
       "release_date": "2026-05-28",
       "last_updated": "2026-05-28",
       "modalities": {
@@ -2102,6 +2398,7 @@
     "global.anthropic.claude-opus-4-7": {
       "id": "global.anthropic.claude-opus-4-7",
       "name": "Claude Opus 4.7 (Global)",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -2147,6 +2444,7 @@
     "us.anthropic.claude-fable-5": {
       "id": "us.anthropic.claude-fable-5",
       "name": "Claude Fable 5 (US)",
+      "description": "Claude model for creative writing, analysis, and controlled agent workflows",
       "family": "claude-fable",
       "attachment": true,
       "reasoning": true,
@@ -2170,7 +2468,8 @@
       "modalities": {
         "input": [
           "text",
-          "image"
+          "image",
+          "pdf"
         ],
         "output": [
           "text"
@@ -2191,6 +2490,7 @@
     "amazon.nova-lite-v1:0": {
       "id": "amazon.nova-lite-v1:0",
       "name": "Nova Lite",
+      "description": "Efficient model for low-latency assistance, extraction, and routine automation",
       "family": "nova-lite",
       "attachment": true,
       "reasoning": false,
@@ -2223,6 +2523,7 @@
     "us.anthropic.claude-haiku-4-5-20251001-v1:0": {
       "id": "us.anthropic.claude-haiku-4-5-20251001-v1:0",
       "name": "Claude Haiku 4.5 (US)",
+      "description": "Fast Claude model for responsive assistance, classification, and lightweight agents",
       "family": "claude-haiku",
       "attachment": true,
       "reasoning": true,
@@ -2263,6 +2564,7 @@
     "xai.grok-4.3": {
       "id": "xai.grok-4.3",
       "name": "Grok 4.3",
+      "description": "xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk",
       "family": "grok",
       "attachment": true,
       "reasoning": true,
@@ -2310,6 +2612,7 @@
     "mistral.voxtral-mini-3b-2507": {
       "id": "mistral.voxtral-mini-3b-2507",
       "name": "Voxtral Mini 3B 2507",
+      "description": "Efficient Mistral model for fast chat, extraction, and production assistants",
       "family": "mistral",
       "attachment": false,
       "reasoning": false,
@@ -2340,6 +2643,7 @@
     "moonshot.kimi-k2-thinking": {
       "id": "moonshot.kimi-k2-thinking",
       "name": "Kimi K2 Thinking",
+      "description": "Kimi reasoning model for long-horizon research, planning, and tool use",
       "family": "kimi-thinking",
       "attachment": false,
       "reasoning": true,
@@ -2371,6 +2675,7 @@
     "meta.llama3-1-70b-instruct-v1:0": {
       "id": "meta.llama3-1-70b-instruct-v1:0",
       "name": "Llama 3.1 70B Instruct",
+      "description": "Open Llama instruction model for multilingual chat, reasoning, and coding",
       "family": "llama",
       "attachment": false,
       "reasoning": false,
@@ -2400,6 +2705,7 @@
     "us.deepseek.r1-v1:0": {
       "id": "us.deepseek.r1-v1:0",
       "name": "DeepSeek-R1 (US)",
+      "description": "Classic open reasoning model for transparent math, coding, and deliberate problem solving",
       "family": "deepseek-thinking",
       "attachment": false,
       "reasoning": true,
@@ -2430,6 +2736,7 @@
     "global.anthropic.claude-sonnet-4-6": {
       "id": "global.anthropic.claude-sonnet-4-6",
       "name": "Claude Sonnet 4.6 (Global)",
+      "description": "Claude workhorse for coding agents, careful analysis, and production cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -2479,6 +2786,7 @@
     "anthropic.claude-haiku-4-5-20251001-v1:0": {
       "id": "anthropic.claude-haiku-4-5-20251001-v1:0",
       "name": "Claude Haiku 4.5",
+      "description": "Fast Claude model for responsive assistance, classification, and lightweight agents",
       "family": "claude-haiku",
       "attachment": true,
       "reasoning": true,
@@ -2519,6 +2827,7 @@
     "moonshotai.kimi-k2.5": {
       "id": "moonshotai.kimi-k2.5",
       "name": "Kimi K2.5",
+      "description": "Kimi multimodal agent model for visual understanding, coding, and planning",
       "family": "kimi",
       "attachment": false,
       "reasoning": true,
@@ -2551,6 +2860,7 @@
     "au.anthropic.claude-opus-4-8": {
       "id": "au.anthropic.claude-opus-4-8",
       "name": "Claude Opus 4.8 (AU)",
+      "description": "Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -2568,6 +2878,7 @@
       ],
       "tool_call": true,
       "temperature": false,
+      "knowledge": "2026-01",
       "release_date": "2026-05-28",
       "last_updated": "2026-05-28",
       "modalities": {
@@ -2595,6 +2906,7 @@
     "nvidia.nemotron-nano-12b-v2": {
       "id": "nvidia.nemotron-nano-12b-v2",
       "name": "NVIDIA Nemotron Nano 12B v2 VL BF16",
+      "description": "Nemotron multimodal model for visual reasoning and agentic AI workflows",
       "family": "nemotron",
       "attachment": false,
       "reasoning": false,
@@ -2625,6 +2937,7 @@
     "zai.glm-4.7-flash": {
       "id": "zai.glm-4.7-flash",
       "name": "GLM-4.7-Flash",
+      "description": "Efficient GLM model for fast reasoning, coding, and agent workflows",
       "family": "glm-flash",
       "attachment": false,
       "reasoning": true,
@@ -2656,6 +2969,7 @@
     "meta.llama4-scout-17b-instruct-v1:0": {
       "id": "meta.llama4-scout-17b-instruct-v1:0",
       "name": "Llama 4 Scout 17B Instruct",
+      "description": "Open multimodal Llama model for long-context analysis and efficient agents",
       "family": "llama",
       "attachment": true,
       "reasoning": false,
@@ -2683,9 +2997,60 @@
         "output": 0.66
       }
     },
+    "au.anthropic.claude-sonnet-5": {
+      "id": "au.anthropic.claude-sonnet-5",
+      "name": "Claude Sonnet 5 (AU)",
+      "description": "Everyday Claude agent model for coding, planning, browsing, and general work",
+      "family": "claude-sonnet",
+      "attachment": true,
+      "reasoning": true,
+      "reasoning_options": [
+        {
+          "type": "toggle"
+        },
+        {
+          "type": "effort",
+          "values": [
+            "low",
+            "medium",
+            "high",
+            "xhigh",
+            "max"
+          ]
+        }
+      ],
+      "tool_call": true,
+      "structured_output": true,
+      "temperature": false,
+      "knowledge": "2026-01-31",
+      "release_date": "2026-06-30",
+      "last_updated": "2026-06-30",
+      "modalities": {
+        "input": [
+          "text",
+          "image",
+          "pdf"
+        ],
+        "output": [
+          "text"
+        ]
+      },
+      "open_weights": false,
+      "limit": {
+        "context": 1000000,
+        "output": 128000
+      },
+      "cost": {
+        "input": 2,
+        "output": 10,
+        "cache_read": 0.2,
+        "cache_write": 2.5
+      }
+    },
     "qwen.qwen3-235b-a22b-2507-v1:0": {
       "id": "qwen.qwen3-235b-a22b-2507-v1:0",
       "name": "Qwen3 235B A22B 2507",
+      "description": "Qwen instruction model for multilingual chat, reasoning, and tool use",
       "family": "qwen",
       "attachment": false,
       "reasoning": false,
@@ -2716,6 +3081,7 @@
     "eu.anthropic.claude-haiku-4-5-20251001-v1:0": {
       "id": "eu.anthropic.claude-haiku-4-5-20251001-v1:0",
       "name": "Claude Haiku 4.5 (EU)",
+      "description": "Fast Claude model for responsive assistance, classification, and lightweight agents",
       "family": "claude-haiku",
       "attachment": true,
       "reasoning": true,
@@ -2756,6 +3122,7 @@
     "openai.gpt-oss-20b-1:0": {
       "id": "openai.gpt-oss-20b-1:0",
       "name": "gpt-oss-20b",
+      "description": "Open-weight GPT model for self-hosted reasoning and instruction-following workloads",
       "family": "gpt-oss",
       "attachment": false,
       "reasoning": true,
@@ -2795,6 +3162,7 @@
     "jp.anthropic.claude-opus-4-8": {
       "id": "jp.anthropic.claude-opus-4-8",
       "name": "Claude Opus 4.8 (JP)",
+      "description": "Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -2812,6 +3180,7 @@
       ],
       "tool_call": true,
       "temperature": false,
+      "knowledge": "2026-01",
       "release_date": "2026-05-28",
       "last_updated": "2026-05-28",
       "modalities": {
@@ -2839,6 +3208,7 @@
     "anthropic.claude-opus-4-8": {
       "id": "anthropic.claude-opus-4-8",
       "name": "Claude Opus 4.8",
+      "description": "Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -2856,6 +3226,7 @@
       ],
       "tool_call": true,
       "temperature": false,
+      "knowledge": "2026-01",
       "release_date": "2026-05-28",
       "last_updated": "2026-05-28",
       "modalities": {
@@ -2883,6 +3254,7 @@
     "qwen.qwen3-coder-30b-a3b-v1:0": {
       "id": "qwen.qwen3-coder-30b-a3b-v1:0",
       "name": "Qwen3 Coder 30B A3B Instruct",
+      "description": "Qwen coding model for software agents, repository edits, and code reasoning",
       "family": "qwen",
       "attachment": false,
       "reasoning": false,
@@ -2913,6 +3285,7 @@
     "qwen.qwen3-next-80b-a3b": {
       "id": "qwen.qwen3-next-80b-a3b",
       "name": "Qwen/Qwen3-Next-80B-A3B-Instruct",
+      "description": "Qwen instruction model for multilingual chat, reasoning, and tool use",
       "family": "qwen",
       "attachment": false,
       "reasoning": false,
@@ -2942,6 +3315,7 @@
     "openai.gpt-5.5": {
       "id": "openai.gpt-5.5",
       "name": "GPT-5.5",
+      "description": "Default frontier GPT for coding, computer use, research, and knowledge work",
       "family": "gpt",
       "attachment": true,
       "reasoning": true,
@@ -2992,6 +3366,7 @@
     "au.anthropic.claude-sonnet-4-6": {
       "id": "au.anthropic.claude-sonnet-4-6",
       "name": "AU Anthropic Claude Sonnet 4.6",
+      "description": "Balanced Claude model for coding, analysis, agent workflows, and cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -3038,9 +3413,60 @@
         "cache_write": 4.125
       }
     },
+    "us.anthropic.claude-sonnet-5": {
+      "id": "us.anthropic.claude-sonnet-5",
+      "name": "Claude Sonnet 5 (US)",
+      "description": "Everyday Claude agent model for coding, planning, browsing, and general work",
+      "family": "claude-sonnet",
+      "attachment": true,
+      "reasoning": true,
+      "reasoning_options": [
+        {
+          "type": "toggle"
+        },
+        {
+          "type": "effort",
+          "values": [
+            "low",
+            "medium",
+            "high",
+            "xhigh",
+            "max"
+          ]
+        }
+      ],
+      "tool_call": true,
+      "structured_output": true,
+      "temperature": false,
+      "knowledge": "2026-01-31",
+      "release_date": "2026-06-30",
+      "last_updated": "2026-06-30",
+      "modalities": {
+        "input": [
+          "text",
+          "image",
+          "pdf"
+        ],
+        "output": [
+          "text"
+        ]
+      },
+      "open_weights": false,
+      "limit": {
+        "context": 1000000,
+        "output": 128000
+      },
+      "cost": {
+        "input": 2,
+        "output": 10,
+        "cache_read": 0.2,
+        "cache_write": 2.5
+      }
+    },
     "us.anthropic.claude-opus-4-5-20251101-v1:0": {
       "id": "us.anthropic.claude-opus-4-5-20251101-v1:0",
       "name": "Claude Opus 4.5 (US)",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -3089,6 +3515,7 @@
     "minimax.minimax-m2.5": {
       "id": "minimax.minimax-m2.5",
       "name": "MiniMax M2.5",
+      "description": "MiniMax model for chat, coding, office work, and agentic tasks",
       "family": "minimax",
       "attachment": false,
       "reasoning": true,
@@ -3119,6 +3546,7 @@
     "eu.anthropic.claude-opus-4-8": {
       "id": "eu.anthropic.claude-opus-4-8",
       "name": "Claude Opus 4.8 (EU)",
+      "description": "Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -3136,6 +3564,7 @@
       ],
       "tool_call": true,
       "temperature": false,
+      "knowledge": "2026-01",
       "release_date": "2026-05-28",
       "last_updated": "2026-05-28",
       "modalities": {
@@ -3163,6 +3592,7 @@
     "eu.anthropic.claude-opus-4-7": {
       "id": "eu.anthropic.claude-opus-4-7",
       "name": "Claude Opus 4.7 (EU)",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -3208,6 +3638,7 @@
     "meta.llama3-1-8b-instruct-v1:0": {
       "id": "meta.llama3-1-8b-instruct-v1:0",
       "name": "Llama 3.1 8B Instruct",
+      "description": "Open Llama instruction model for multilingual chat, reasoning, and coding",
       "family": "llama",
       "attachment": false,
       "reasoning": false,
@@ -3237,6 +3668,7 @@
     "us.anthropic.claude-opus-4-1-20250805-v1:0": {
       "id": "us.anthropic.claude-opus-4-1-20250805-v1:0",
       "name": "Claude Opus 4.1 (US)",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -3276,6 +3708,7 @@
     "meta.llama4-maverick-17b-instruct-v1:0": {
       "id": "meta.llama4-maverick-17b-instruct-v1:0",
       "name": "Llama 4 Maverick 17B Instruct",
+      "description": "Open multimodal Llama model for strong reasoning and fast responses",
       "family": "llama",
       "attachment": true,
       "reasoning": false,
@@ -3306,6 +3739,7 @@
     "global.anthropic.claude-opus-4-5-20251101-v1:0": {
       "id": "global.anthropic.claude-opus-4-5-20251101-v1:0",
       "name": "Claude Opus 4.5 (Global)",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -3354,6 +3788,7 @@
     "nvidia.nemotron-super-3-120b": {
       "id": "nvidia.nemotron-super-3-120b",
       "name": "NVIDIA Nemotron 3 Super 120B A12B",
+      "description": "Nemotron middle tier for collaborative agents and high-volume reasoning workloads",
       "family": "nemotron",
       "attachment": false,
       "reasoning": true,
@@ -3381,9 +3816,60 @@
         "output": 0.65
       }
     },
+    "eu.anthropic.claude-sonnet-5": {
+      "id": "eu.anthropic.claude-sonnet-5",
+      "name": "Claude Sonnet 5 (EU)",
+      "description": "Everyday Claude agent model for coding, planning, browsing, and general work",
+      "family": "claude-sonnet",
+      "attachment": true,
+      "reasoning": true,
+      "reasoning_options": [
+        {
+          "type": "toggle"
+        },
+        {
+          "type": "effort",
+          "values": [
+            "low",
+            "medium",
+            "high",
+            "xhigh",
+            "max"
+          ]
+        }
+      ],
+      "tool_call": true,
+      "structured_output": true,
+      "temperature": false,
+      "knowledge": "2026-01-31",
+      "release_date": "2026-06-30",
+      "last_updated": "2026-06-30",
+      "modalities": {
+        "input": [
+          "text",
+          "image",
+          "pdf"
+        ],
+        "output": [
+          "text"
+        ]
+      },
+      "open_weights": false,
+      "limit": {
+        "context": 1000000,
+        "output": 128000
+      },
+      "cost": {
+        "input": 2.2,
+        "output": 11,
+        "cache_read": 0.22,
+        "cache_write": 2.75
+      }
+    },
     "nvidia.nemotron-nano-3-30b": {
       "id": "nvidia.nemotron-nano-3-30b",
       "name": "NVIDIA Nemotron Nano 3 30B",
+      "description": "Small Nemotron 3 MoE for efficient coding, math, and long-context agents",
       "family": "nemotron",
       "attachment": false,
       "reasoning": true,
@@ -3414,6 +3900,7 @@
     "eu.anthropic.claude-opus-4-5-20251101-v1:0": {
       "id": "eu.anthropic.claude-opus-4-5-20251101-v1:0",
       "name": "Claude Opus 4.5 (EU)",
+      "description": "Flagship Claude model for deep reasoning, coding, and long-horizon agents",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -3462,6 +3949,7 @@
     "global.anthropic.claude-opus-4-6-v1": {
       "id": "global.anthropic.claude-opus-4-6-v1",
       "name": "Claude Opus 4.6 (Global)",
+      "description": "High-end Claude for difficult coding, planning, and slower expert reasoning",
       "family": "claude-opus",
       "attachment": true,
       "reasoning": true,
@@ -3511,6 +3999,7 @@
     "us.anthropic.claude-sonnet-4-5-20250929-v1:0": {
       "id": "us.anthropic.claude-sonnet-4-5-20250929-v1:0",
       "name": "Claude Sonnet 4.5 (US)",
+      "description": "Balanced Claude model for coding, analysis, agent workflows, and cost control",
       "family": "claude-sonnet",
       "attachment": true,
       "reasoning": true,
@@ -3551,6 +4040,7 @@
     "google.gemma-3-12b-it": {
       "id": "google.gemma-3-12b-it",
       "name": "Google Gemma 3 12B",
+      "description": "Open Gemma instruction model for efficient chat and self-hosted deployments",
       "family": "gemma",
       "attachment": false,
       "reasoning": false,
@@ -3582,6 +4072,7 @@
     "minimax.minimax-m2.1": {
       "id": "minimax.minimax-m2.1",
       "name": "MiniMax M2.1",
+      "description": "MiniMax model for chat, coding, office work, and agentic tasks",
       "family": "minimax",
       "attachment": false,
       "reasoning": true,
@@ -3612,6 +4103,7 @@
     "eu.anthropic.claude-fable-5": {
       "id": "eu.anthropic.claude-fable-5",
       "name": "Claude Fable 5 (EU)",
+      "description": "Claude model for creative writing, analysis, and controlled agent workflows",
       "family": "claude-fable",
       "attachment": true,
       "reasoning": true,
@@ -3635,7 +4127,8 @@
       "modalities": {
         "input": [
           "text",
-          "image"
+          "image",
+          "pdf"
         ],
         "output": [
           "text"
@@ -3656,6 +4149,7 @@
     "deepseek.v3.2": {
       "id": "deepseek.v3.2",
       "name": "DeepSeek-V3.2",
+      "description": "DeepSeek chat model for instruction following, coding, and analysis",
       "family": "deepseek",
       "attachment": false,
       "reasoning": true,
@@ -3687,6 +4181,7 @@
     "mistral.ministral-3-14b-instruct": {
       "id": "mistral.ministral-3-14b-instruct",
       "name": "Ministral 14B 3.0",
+      "description": "Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads",
       "family": "ministral",
       "attachment": false,
       "reasoning": false,