llm.rb 4.14.0 → 4.16.0

data/data/zai.json

@@ -8,37 +8,6 @@
   "name": "Z.AI",
   "doc": "https://docs.z.ai/guides/overview/pricing",
   "models": {
-    "glm-4.7-flash": {
-      "id": "glm-4.7-flash",
-      "name": "GLM-4.7-Flash",
-      "family": "glm-flash",
-      "attachment": false,
-      "reasoning": true,
-      "tool_call": true,
-      "temperature": true,
-      "knowledge": "2025-04",
-      "release_date": "2026-01-19",
-      "last_updated": "2026-01-19",
-      "modalities": {
-        "input": [
-          "text"
-        ],
-        "output": [
-          "text"
-        ]
-      },
-      "open_weights": true,
-      "cost": {
-        "input": 0,
-        "output": 0,
-        "cache_read": 0,
-        "cache_write": 0
-      },
-      "limit": {
-        "context": 200000,
-        "output": 131072
-      }
-    },
     "glm-5v-turbo": {
       "id": "glm-5v-turbo",
       "name": "glm-5v-turbo",
@@ -75,9 +44,9 @@
         "output": 131072
       }
     },
-    "glm-5-turbo": {
-      "id": "glm-5-turbo",
-      "name": "GLM-5-Turbo",
+    "glm-4.7": {
+      "id": "glm-4.7",
+      "name": "GLM-4.7",
       "family": "glm",
       "attachment": false,
       "reasoning": true,
@@ -85,10 +54,10 @@
       "interleaved": {
         "field": "reasoning_content"
       },
-      "structured_output": true,
       "temperature": true,
-      "release_date": "2026-03-16",
-      "last_updated": "2026-03-16",
+      "knowledge": "2025-04",
+      "release_date": "2025-12-22",
+      "last_updated": "2025-12-22",
       "modalities": {
         "input": [
           "text"
@@ -97,29 +66,31 @@
           "text"
         ]
       },
-      "open_weights": false,
+      "open_weights": true,
       "cost": {
-        "input": 1.2,
-        "output": 4,
-        "cache_read": 0.24,
+        "input": 0.6,
+        "output": 2.2,
+        "cache_read": 0.11,
         "cache_write": 0
       },
       "limit": {
-        "context": 200000,
+        "context": 204800,
         "output": 131072
       }
     },
-    "glm-4.5": {
-      "id": "glm-4.5",
-      "name": "GLM-4.5",
+    "glm-5": {
+      "id": "glm-5",
+      "name": "GLM-5",
       "family": "glm",
       "attachment": false,
       "reasoning": true,
       "tool_call": true,
+      "interleaved": {
+        "field": "reasoning_content"
+      },
       "temperature": true,
-      "knowledge": "2025-04",
-      "release_date": "2025-07-28",
-      "last_updated": "2025-07-28",
+      "release_date": "2026-02-11",
+      "last_updated": "2026-02-11",
       "modalities": {
         "input": [
           "text"
@@ -130,14 +101,14 @@
       },
       "open_weights": true,
       "cost": {
-        "input": 0.6,
-        "output": 2.2,
-        "cache_read": 0.11,
+        "input": 1,
+        "output": 3.2,
+        "cache_read": 0.2,
         "cache_write": 0
       },
       "limit": {
-        "context": 131072,
-        "output": 98304
+        "context": 204800,
+        "output": 131072
       }
     },
     "glm-4.7-flashx": {
@@ -171,17 +142,20 @@
         "output": 131072
       }
     },
-    "glm-4.6": {
-      "id": "glm-4.6",
-      "name": "GLM-4.6",
+    "glm-5.1": {
+      "id": "glm-5.1",
+      "name": "GLM-5.1",
       "family": "glm",
       "attachment": false,
       "reasoning": true,
       "tool_call": true,
+      "interleaved": {
+        "field": "reasoning_content"
+      },
+      "structured_output": true,
       "temperature": true,
-      "knowledge": "2025-04",
-      "release_date": "2025-09-30",
-      "last_updated": "2025-09-30",
+      "release_date": "2026-03-27",
+      "last_updated": "2026-03-27",
       "modalities": {
         "input": [
           "text"
@@ -190,34 +164,32 @@
           "text"
         ]
       },
-      "open_weights": true,
+      "open_weights": false,
       "cost": {
-        "input": 0.6,
-        "output": 2.2,
-        "cache_read": 0.11,
+        "input": 1.4,
+        "output": 4.4,
+        "cache_read": 0.26,
         "cache_write": 0
       },
       "limit": {
-        "context": 204800,
+        "context": 200000,
         "output": 131072
       }
     },
-    "glm-4.6v": {
-      "id": "glm-4.6v",
-      "name": "GLM-4.6V",
+    "glm-4.5": {
+      "id": "glm-4.5",
+      "name": "GLM-4.5",
       "family": "glm",
-      "attachment": true,
+      "attachment": false,
       "reasoning": true,
       "tool_call": true,
       "temperature": true,
       "knowledge": "2025-04",
-      "release_date": "2025-12-08",
-      "last_updated": "2025-12-08",
+      "release_date": "2025-07-28",
+      "last_updated": "2025-07-28",
       "modalities": {
         "input": [
-          "text",
-          "image",
-          "video"
+          "text"
         ],
         "output": [
           "text"
@@ -225,18 +197,20 @@
       },
       "open_weights": true,
       "cost": {
-        "input": 0.3,
-        "output": 0.9
+        "input": 0.6,
+        "output": 2.2,
+        "cache_read": 0.11,
+        "cache_write": 0
       },
       "limit": {
-        "context": 128000,
-        "output": 32768
+        "context": 131072,
+        "output": 98304
       }
     },
-    "glm-4.5-flash": {
-      "id": "glm-4.5-flash",
-      "name": "GLM-4.5-Flash",
-      "family": "glm-flash",
+    "glm-4.5-air": {
+      "id": "glm-4.5-air",
+      "name": "GLM-4.5-Air",
+      "family": "glm-air",
       "attachment": false,
       "reasoning": true,
       "tool_call": true,
@@ -254,9 +228,9 @@
       },
       "open_weights": true,
       "cost": {
-        "input": 0,
-        "output": 0,
-        "cache_read": 0,
+        "input": 0.2,
+        "output": 1.1,
+        "cache_read": 0.03,
         "cache_write": 0
       },
       "limit": {
@@ -264,9 +238,9 @@
         "output": 98304
       }
     },
-    "glm-5": {
-      "id": "glm-5",
-      "name": "GLM-5",
+    "glm-5-turbo": {
+      "id": "glm-5-turbo",
+      "name": "GLM-5-Turbo",
       "family": "glm",
       "attachment": false,
       "reasoning": true,
@@ -274,9 +248,10 @@
       "interleaved": {
         "field": "reasoning_content"
       },
+      "structured_output": true,
       "temperature": true,
-      "release_date": "2026-02-11",
-      "last_updated": "2026-02-11",
+      "release_date": "2026-03-16",
+      "last_updated": "2026-03-16",
       "modalities": {
         "input": [
           "text"
@@ -285,29 +260,60 @@
           "text"
         ]
       },
-      "open_weights": true,
+      "open_weights": false,
       "cost": {
-        "input": 1,
-        "output": 3.2,
-        "cache_read": 0.2,
+        "input": 1.2,
+        "output": 4,
+        "cache_read": 0.24,
         "cache_write": 0
       },
       "limit": {
-        "context": 204800,
+        "context": 200000,
         "output": 131072
       }
     },
-    "glm-4.5-air": {
-      "id": "glm-4.5-air",
-      "name": "GLM-4.5-Air",
-      "family": "glm-air",
+    "glm-4.5v": {
+      "id": "glm-4.5v",
+      "name": "GLM-4.5V",
+      "family": "glm",
+      "attachment": true,
+      "reasoning": true,
+      "tool_call": true,
+      "temperature": true,
+      "knowledge": "2025-04",
+      "release_date": "2025-08-11",
+      "last_updated": "2025-08-11",
+      "modalities": {
+        "input": [
+          "text",
+          "image",
+          "video"
+        ],
+        "output": [
+          "text"
+        ]
+      },
+      "open_weights": true,
+      "cost": {
+        "input": 0.6,
+        "output": 1.8
+      },
+      "limit": {
+        "context": 64000,
+        "output": 16384
+      }
+    },
+    "glm-4.6": {
+      "id": "glm-4.6",
+      "name": "GLM-4.6",
+      "family": "glm",
       "attachment": false,
       "reasoning": true,
       "tool_call": true,
       "temperature": true,
       "knowledge": "2025-04",
-      "release_date": "2025-07-28",
-      "last_updated": "2025-07-28",
+      "release_date": "2025-09-30",
+      "last_updated": "2025-09-30",
       "modalities": {
         "input": [
           "text"
@@ -318,27 +324,27 @@
       },
       "open_weights": true,
       "cost": {
-        "input": 0.2,
-        "output": 1.1,
-        "cache_read": 0.03,
+        "input": 0.6,
+        "output": 2.2,
+        "cache_read": 0.11,
         "cache_write": 0
       },
       "limit": {
-        "context": 131072,
-        "output": 98304
+        "context": 204800,
+        "output": 131072
       }
     },
-    "glm-4.5v": {
-      "id": "glm-4.5v",
-      "name": "GLM-4.5V",
+    "glm-4.6v": {
+      "id": "glm-4.6v",
+      "name": "GLM-4.6V",
       "family": "glm",
       "attachment": true,
       "reasoning": true,
       "tool_call": true,
       "temperature": true,
       "knowledge": "2025-04",
-      "release_date": "2025-08-11",
-      "last_updated": "2025-08-11",
+      "release_date": "2025-12-08",
+      "last_updated": "2025-12-08",
       "modalities": {
         "input": [
           "text",
@@ -351,28 +357,56 @@
       },
       "open_weights": true,
       "cost": {
-        "input": 0.6,
-        "output": 1.8
+        "input": 0.3,
+        "output": 0.9
       },
       "limit": {
-        "context": 64000,
-        "output": 16384
+        "context": 128000,
+        "output": 32768
       }
     },
-    "glm-4.7": {
-      "id": "glm-4.7",
-      "name": "GLM-4.7",
-      "family": "glm",
+    "glm-4.5-flash": {
+      "id": "glm-4.5-flash",
+      "name": "GLM-4.5-Flash",
+      "family": "glm-flash",
       "attachment": false,
       "reasoning": true,
       "tool_call": true,
-      "interleaved": {
-        "field": "reasoning_content"
+      "temperature": true,
+      "knowledge": "2025-04",
+      "release_date": "2025-07-28",
+      "last_updated": "2025-07-28",
+      "modalities": {
+        "input": [
+          "text"
+        ],
+        "output": [
+          "text"
+        ]
       },
+      "open_weights": true,
+      "cost": {
+        "input": 0,
+        "output": 0,
+        "cache_read": 0,
+        "cache_write": 0
+      },
+      "limit": {
+        "context": 131072,
+        "output": 98304
+      }
+    },
+    "glm-4.7-flash": {
+      "id": "glm-4.7-flash",
+      "name": "GLM-4.7-Flash",
+      "family": "glm-flash",
+      "attachment": false,
+      "reasoning": true,
+      "tool_call": true,
       "temperature": true,
       "knowledge": "2025-04",
-      "release_date": "2025-12-22",
-      "last_updated": "2025-12-22",
+      "release_date": "2026-01-19",
+      "last_updated": "2026-01-19",
       "modalities": {
         "input": [
           "text"
@@ -383,13 +417,13 @@
       },
       "open_weights": true,
       "cost": {
-        "input": 0.6,
-        "output": 2.2,
-        "cache_read": 0.11,
+        "input": 0,
+        "output": 0,
+        "cache_read": 0,
         "cache_write": 0
       },
       "limit": {
-        "context": 204800,
+        "context": 200000,
         "output": 131072
       }
     }

data/lib/llm/active_record/acts_as_llm.rb

@@ -0,0 +1,238 @@
+# frozen_string_literal: true
+
+module LLM::ActiveRecord
+  ##
+  # ActiveRecord integration for persisting {LLM::Context LLM::Context} state.
+  #
+  # This wrapper maps model columns onto provider selection, model selection,
+  # usage accounting, and serialized context data while leaving application-
+  # specific concerns such as credentials, associations, and UI shaping to
+  # the host app.
+  #
+  # Context state can be stored as a JSON string (`format: :string`, the
+  # default) or as a structured object (`format: :json` / `:jsonb`) for
+  # databases such as PostgreSQL that can persist JSON natively.
+  # `:json` and `:jsonb` expect a real JSON column type with ActiveRecord
+  # handling JSON typecasting for the model.
+  module ActsAsLLM
+    EMPTY_HASH = {}.freeze
+    DEFAULT_USAGE_COLUMNS = {
+      input_tokens: :input_tokens,
+      output_tokens: :output_tokens,
+      total_tokens: :total_tokens
+    }.freeze
+    DEFAULTS = {
+      provider_column: :provider,
+      model_column: :model,
+      data_column: :data,
+      format: :string,
+      usage_columns: DEFAULT_USAGE_COLUMNS,
+      provider: EMPTY_HASH,
+      context: EMPTY_HASH
+    }.freeze
+
+    module Hooks
+      ##
+      # Called when hooks are extended onto an ActiveRecord model.
+      #
+      # @param [Class] model
+      # @return [void]
+      def self.extended(model)
+        options = model.llm_plugin_options
+        model.validates options[:provider_column], options[:model_column], presence: true
+        model.include InstanceMethods unless model.ancestors.include?(InstanceMethods)
+      end
+    end
+
+    ##
+    # Installs the `acts_as_llm` wrapper on an ActiveRecord model.
+    #
+    # @param [Hash] options
+    # @option options [Symbol] :format
+    #   Storage format for the serialized context. Use `:string` for text
+    #   columns, or `:json` / `:jsonb` for structured JSON columns with
+    #   ActiveRecord JSON typecasting enabled.
+    # @return [void]
+    def acts_as_llm(options = EMPTY_HASH)
+      options = DEFAULTS.merge(options)
+      usage_columns = DEFAULT_USAGE_COLUMNS.merge(options[:usage_columns] || EMPTY_HASH)
+      class_attribute :llm_plugin_options, instance_accessor: false, default: DEFAULTS unless respond_to?(:llm_plugin_options)
+      self.llm_plugin_options = options.merge(usage_columns: usage_columns.freeze).freeze
+      extend Hooks
+    end
+
+    module InstanceMethods
+      ##
+      # Continues the stored context with new input and flushes it.
+      # @see LLM::Context#talk
+      # @return [LLM::Response]
+      def talk(...)
+        ctx.talk(...).tap { flush }
+      end
+
+      ##
+      # Continues the stored context through the Responses API and flushes it.
+      # @see LLM::Context#respond
+      # @return [LLM::Response]
+      def respond(...)
+        ctx.respond(...).tap { flush }
+      end
+
+      ##
+      # Waits for queued tool work to finish.
+      # @see LLM::Context#wait
+      # @return [Array<LLM::Function::Return>]
+      def wait(...)
+        ctx.wait(...)
+      end
+
+      ##
+      # Calls into the stored context.
+      # @see LLM::Context#call
+      # @return [Object]
+      def call(...)
+        ctx.call(...)
+      end
+
+      ##
+      # @see LLM::Context#messages
+      # @return [Array<LLM::Message>]
+      def messages
+        ctx.messages
+      end
+
+      ##
+      # @note The bang keeps the ActiveRecord and Sequel wrappers aligned.
+      # @see LLM::Context#model
+      # @return [String]
+      def model!
+        ctx.model
+      end
+
+      ##
+      # @see LLM::Context#functions
+      # @return [Array<LLM::Function>]
+      def functions
+        ctx.functions
+      end
+
+      ##
+      # @see LLM::Context#cost
+      # @return [LLM::Cost]
+      def cost
+        ctx.cost
+      end
+
+      ##
+      # @see LLM::Context#context_window
+      # @return [Integer]
+      def context_window
+        ctx.context_window
+      rescue LLM::NoSuchModelError, LLM::NoSuchRegistryError
+        0
+      end
+
+      ##
+      # Returns usage from the mapped usage columns.
+      # @return [LLM::Object]
+      def usage
+        LLM::Object.from(
+          input_tokens: self[columns[:input_tokens]] || 0,
+          output_tokens: self[columns[:output_tokens]] || 0,
+          total_tokens: self[columns[:total_tokens]] || 0
+        )
+      end
+
+      private
+
+      ##
+      # Returns the resolved provider instance for this record.
+      # @return [LLM::Provider]
+      def llm
+        options = self.class.llm_plugin_options
+        provider = self[columns[:provider_column]]
+        kwargs = resolve_options(options[:provider])
+        @llm ||= LLM.method(provider).call(**kwargs)
+      end
+
+      ##
+      # @return [LLM::Context]
+      def ctx
+        @ctx ||= begin
+          options = self.class.llm_plugin_options
+          params = resolve_options(options[:context]).dup
+          params[:model] ||= self[columns[:model_column]]
+          ctx = LLM::Context.new(llm, params.compact)
+          data = self[columns[:data_column]]
+          if data.nil? || data == ""
+            ctx
+          else
+            string = case options[:format]
+            when :string then data
+            when :json, :jsonb then LLM.json.dump(data)
+            else raise ArgumentError, "Unknown format: #{options[:format].inspect}"
+            end
+            ctx.restore(string:)
+          end
+        end
+      end
+
+      ##
+      # @return [void]
+      def flush
+        attrs = {
+          columns[:data_column] => serialize_context(self.class.llm_plugin_options[:format]),
+          columns[:input_tokens] => ctx.usage.input_tokens,
+          columns[:output_tokens] => ctx.usage.output_tokens,
+          columns[:total_tokens] => ctx.usage.total_tokens
+        }
+        assign_attributes(attrs)
+        save!
+      end
+
+      ##
+      # @return [Hash]
+      def resolve_option(option)
+        case option
+        when Proc then instance_exec(&option)
+        when Hash then option.dup
+        else option
+        end
+      end
+
+      ##
+      # @return [Hash]
+      def resolve_options(option)
+        case option
+        when Proc, Hash then resolve_option(option)
+        else EMPTY_HASH.dup
+        end
+      end
+
+      def serialize_context(format)
+        case format
+        when :string then ctx.to_json
+        when :json, :jsonb then ctx.to_h
+        else raise ArgumentError, "Unknown format: #{format.inspect}"
+        end
+      end
+
+      def columns
+        @columns ||= begin
+          options = self.class.llm_plugin_options
+          usage_columns = options[:usage_columns]
+          {
+            provider_column: options[:provider_column],
+            model_column: options[:model_column],
+            data_column: options[:data_column],
+            input_tokens: usage_columns[:input_tokens],
+            output_tokens: usage_columns[:output_tokens],
+            total_tokens: usage_columns[:total_tokens]
+          }.freeze
+        end
+      end
+    end
+  end
+end
+
+::ActiveRecord::Base.extend(LLM::ActiveRecord::ActsAsLLM)