ruby_llm 1.10.0 → 1.12.0

data/lib/ruby_llm/providers/xai/models.rb

@@ -0,0 +1,75 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Providers
+    class XAI
+      # Models metadata for xAI list models.
+      module Models
+        module_function
+
+        IMAGE_MODELS = %w[grok-2-image-1212].freeze
+        VISION_MODELS = %w[
+          grok-2-vision-1212
+          grok-4-0709
+          grok-4-fast-non-reasoning
+          grok-4-fast-reasoning
+          grok-4-1-fast-non-reasoning
+          grok-4-1-fast-reasoning
+        ].freeze
+        REASONING_MODELS = %w[
+          grok-3-mini
+          grok-4-0709
+          grok-4-fast-reasoning
+          grok-4-1-fast-reasoning
+          grok-code-fast-1
+        ].freeze
+
+        def parse_list_models_response(response, slug, _capabilities)
+          Array(response.body['data']).map do |model_data|
+            model_id = model_data['id']
+
+            Model::Info.new(
+              id: model_id,
+              name: format_display_name(model_id),
+              provider: slug,
+              family: 'grok',
+              created_at: model_data['created'] ? Time.at(model_data['created']) : nil,
+              context_window: nil,
+              max_output_tokens: nil,
+              modalities: modalities_for(model_id),
+              capabilities: capabilities_for(model_id),
+              pricing: {},
+              metadata: {
+                object: model_data['object'],
+                owned_by: model_data['owned_by']
+              }.compact
+            )
+          end
+        end
+
+        def modalities_for(model_id)
+          if IMAGE_MODELS.include?(model_id)
+            { input: ['text'], output: ['image'] }
+          else
+            input = ['text']
+            input << 'image' if VISION_MODELS.include?(model_id)
+            { input: input, output: ['text'] }
+          end
+        end
+
+        def capabilities_for(model_id)
+          return [] if IMAGE_MODELS.include?(model_id)
+
+          capabilities = %w[streaming function_calling structured_output]
+          capabilities << 'reasoning' if REASONING_MODELS.include?(model_id)
+          capabilities << 'vision' if VISION_MODELS.include?(model_id)
+          capabilities
+        end
+
+        def format_display_name(model_id)
+          model_id.tr('-', ' ').split.map(&:capitalize).join(' ')
+        end
+      end
+    end
+  end
+end

data/lib/ruby_llm/providers/xai.rb

@@ -0,0 +1,28 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Providers
+    # xAI API integration
+    class XAI < OpenAI
+      include XAI::Chat
+      include XAI::Models
+
+      def api_base
+        'https://api.x.ai/v1'
+      end
+
+      def headers
+        {
+          'Authorization' => "Bearer #{@config.xai_api_key}",
+          'Content-Type' => 'application/json'
+        }
+      end
+
+      class << self
+        def configuration_requirements
+          %i[xai_api_key]
+        end
+      end
+    end
+  end
+end

data/lib/ruby_llm/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module RubyLLM
-  VERSION = '1.10.0'
+  VERSION = '1.12.0'
 end

data/lib/ruby_llm.rb

@@ -9,23 +9,27 @@ require 'json'
 require 'logger'
 require 'marcel'
 require 'securerandom'
+require 'date'
+require 'time'
 require 'zeitwerk'
 
 loader = Zeitwerk::Loader.for_gem
 loader.inflector.inflect(
-  'ruby_llm' => 'RubyLLM',
-  'llm' => 'LLM',
-  'openai' => 'OpenAI',
+  'azure' => 'Azure',
+  'UI' => 'UI',
   'api' => 'API',
-  'deepseek' => 'DeepSeek',
-  'perplexity' => 'Perplexity',
   'bedrock' => 'Bedrock',
-  'openrouter' => 'OpenRouter',
+  'deepseek' => 'DeepSeek',
   'gpustack' => 'GPUStack',
+  'llm' => 'LLM',
   'mistral' => 'Mistral',
-  'vertexai' => 'VertexAI',
+  'openai' => 'OpenAI',
+  'openrouter' => 'OpenRouter',
   'pdf' => 'PDF',
-  'UI' => 'UI'
+  'perplexity' => 'Perplexity',
+  'ruby_llm' => 'RubyLLM',
+  'vertexai' => 'VertexAI',
+  'xai' => 'XAI'
 )
 loader.ignore("#{__dir__}/tasks")
 loader.ignore("#{__dir__}/generators")
@@ -90,6 +94,7 @@ module RubyLLM
 end
 
 RubyLLM::Provider.register :anthropic, RubyLLM::Providers::Anthropic
+RubyLLM::Provider.register :azure, RubyLLM::Providers::Azure
 RubyLLM::Provider.register :bedrock, RubyLLM::Providers::Bedrock
 RubyLLM::Provider.register :deepseek, RubyLLM::Providers::DeepSeek
 RubyLLM::Provider.register :gemini, RubyLLM::Providers::Gemini
@@ -100,6 +105,7 @@ RubyLLM::Provider.register :openai, RubyLLM::Providers::OpenAI
 RubyLLM::Provider.register :openrouter, RubyLLM::Providers::OpenRouter
 RubyLLM::Provider.register :perplexity, RubyLLM::Providers::Perplexity
 RubyLLM::Provider.register :vertexai, RubyLLM::Providers::VertexAI
+RubyLLM::Provider.register :xai, RubyLLM::Providers::XAI
 
 if defined?(Rails::Railtie)
   require 'ruby_llm/railtie'

data/lib/tasks/models.rake

@@ -39,15 +39,18 @@ end
 
 def configure_from_env
   RubyLLM.configure do |config|
-    config.openai_api_key = ENV.fetch('OPENAI_API_KEY', nil)
     config.anthropic_api_key = ENV.fetch('ANTHROPIC_API_KEY', nil)
-    config.gemini_api_key = ENV.fetch('GEMINI_API_KEY', nil)
+    config.azure_api_base = ENV.fetch('AZURE_API_BASE', nil)
+    config.azure_api_key = ENV.fetch('AZURE_API_KEY', nil)
     config.deepseek_api_key = ENV.fetch('DEEPSEEK_API_KEY', nil)
-    config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', nil)
-    config.openrouter_api_key = ENV.fetch('OPENROUTER_API_KEY', nil)
+    config.gemini_api_key = ENV.fetch('GEMINI_API_KEY', nil)
     config.mistral_api_key = ENV.fetch('MISTRAL_API_KEY', nil)
+    config.openai_api_key = ENV.fetch('OPENAI_API_KEY', nil)
+    config.openrouter_api_key = ENV.fetch('OPENROUTER_API_KEY', nil)
+    config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', nil)
     config.vertexai_location = ENV.fetch('GOOGLE_CLOUD_LOCATION', nil)
     config.vertexai_project_id = ENV.fetch('GOOGLE_CLOUD_PROJECT', nil)
+    config.xai_api_key = ENV.fetch('XAI_API_KEY', nil)
     configure_bedrock(config)
     config.request_timeout = 30
   end
@@ -330,6 +333,7 @@ def generate_aliases # rubocop:disable Metrics/PerceivedComplexity
   # OpenAI models
   models['openai'].each do |model|
     openrouter_model = "openai/#{model}"
+    azure_model = models['azure'].include?(model) ? model : nil
     next unless models['openrouter'].include?(openrouter_model)
 
     alias_key = model.gsub('-latest', '')
@@ -337,6 +341,7 @@ def generate_aliases # rubocop:disable Metrics/PerceivedComplexity
       'openai' => model,
       'openrouter' => openrouter_model
     }
+    aliases[alias_key]['azure'] = azure_model if azure_model
   end
 
   anthropic_latest = group_anthropic_models_by_base_name(models['anthropic'])
@@ -357,6 +362,7 @@ def generate_aliases # rubocop:disable Metrics/PerceivedComplexity
     aliases[base_name] = { 'anthropic' => latest_model }
     aliases[base_name]['openrouter'] = openrouter_model if openrouter_model
     aliases[base_name]['bedrock'] = bedrock_model if bedrock_model
+    aliases[base_name]['azure'] = latest_model if models['azure'].include?(latest_model)
   end
 
   models['bedrock'].each do |bedrock_model|

data/lib/tasks/vcr.rake

@@ -1,6 +1,7 @@
 # frozen_string_literal: true
 
 require 'dotenv/load'
+require 'time'
 
 def record_all_cassettes(cassette_dir)
   FileUtils.rm_rf(cassette_dir)
@@ -73,6 +74,31 @@ def run_tests
   system('bundle exec rspec') || abort('Tests failed')
 end
 
+def retimestamp_cassettes(cassette_dir)
+  timestamp = Time.now.utc.httpdate
+  updated_files = 0
+  updated_entries = 0
+
+  Dir.glob("#{cassette_dir}/**/*.yml").each do |file|
+    content = File.read(file)
+    replacements = 0
+
+    updated = content.gsub(/^(\s*recorded_at:\s*).+$/) do
+      replacements += 1
+      "#{Regexp.last_match(1)}#{timestamp}"
+    end
+
+    next if replacements.zero?
+
+    File.write(file, updated)
+    updated_files += 1
+    updated_entries += replacements
+  end
+
+  puts "Updated #{updated_entries} recorded_at entries in #{updated_files} cassette files."
+  puts "New recorded_at value: #{timestamp}"
+end
+
 namespace :vcr do
   desc 'Record VCR cassettes (rake vcr:record[all] or vcr:record[openai,anthropic])'
   task :record, :providers do |_, args|
@@ -89,4 +115,10 @@ namespace :vcr do
       record_for_providers(providers, cassette_dir)
     end
   end
+
+  desc 'Update recorded_at timestamps for all cassette entries'
+  task :retimestamp do
+    cassette_dir = 'spec/fixtures/vcr_cassettes'
+    retimestamp_cassettes(cassette_dir)
+  end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: ruby_llm
 version: !ruby/object:Gem::Version
-  version: 1.10.0
+  version: 1.12.0
 platform: ruby
 authors:
 - Carmine Paolino
@@ -43,28 +43,28 @@ dependencies:
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 1.10.0
+        version: '2.0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 1.10.0
+        version: '2.0'
 - !ruby/object:Gem::Dependency
   name: faraday-retry
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: '1'
+        version: '2.0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: '1'
+        version: '2.0'
 - !ruby/object:Gem::Dependency
   name: faraday-multipart
   requirement: !ruby/object:Gem::Requirement
@@ -192,6 +192,7 @@ files:
 - lib/ruby_llm/active_record/chat_methods.rb
 - lib/ruby_llm/active_record/message_methods.rb
 - lib/ruby_llm/active_record/model_methods.rb
+- lib/ruby_llm/agent.rb
 - lib/ruby_llm/aliases.json
 - lib/ruby_llm/aliases.rb
 - lib/ruby_llm/attachment.rb
@@ -226,18 +227,17 @@ files:
 - lib/ruby_llm/providers/anthropic/models.rb
 - lib/ruby_llm/providers/anthropic/streaming.rb
 - lib/ruby_llm/providers/anthropic/tools.rb
+- lib/ruby_llm/providers/azure.rb
+- lib/ruby_llm/providers/azure/chat.rb
+- lib/ruby_llm/providers/azure/embeddings.rb
+- lib/ruby_llm/providers/azure/media.rb
+- lib/ruby_llm/providers/azure/models.rb
 - lib/ruby_llm/providers/bedrock.rb
-- lib/ruby_llm/providers/bedrock/capabilities.rb
+- lib/ruby_llm/providers/bedrock/auth.rb
 - lib/ruby_llm/providers/bedrock/chat.rb
 - lib/ruby_llm/providers/bedrock/media.rb
 - lib/ruby_llm/providers/bedrock/models.rb
-- lib/ruby_llm/providers/bedrock/signing.rb
 - lib/ruby_llm/providers/bedrock/streaming.rb
-- lib/ruby_llm/providers/bedrock/streaming/base.rb
-- lib/ruby_llm/providers/bedrock/streaming/content_extraction.rb
-- lib/ruby_llm/providers/bedrock/streaming/message_processing.rb
-- lib/ruby_llm/providers/bedrock/streaming/payload_processing.rb
-- lib/ruby_llm/providers/bedrock/streaming/prelude_handling.rb
 - lib/ruby_llm/providers/deepseek.rb
 - lib/ruby_llm/providers/deepseek/capabilities.rb
 - lib/ruby_llm/providers/deepseek/chat.rb
@@ -290,6 +290,9 @@ files:
 - lib/ruby_llm/providers/vertexai/models.rb
 - lib/ruby_llm/providers/vertexai/streaming.rb
 - lib/ruby_llm/providers/vertexai/transcription.rb
+- lib/ruby_llm/providers/xai.rb
+- lib/ruby_llm/providers/xai/chat.rb
+- lib/ruby_llm/providers/xai/models.rb
 - lib/ruby_llm/railtie.rb
 - lib/ruby_llm/stream_accumulator.rb
 - lib/ruby_llm/streaming.rb
@@ -332,7 +335,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.6.9
+rubygems_version: 4.0.3
 specification_version: 4
 summary: One beautiful Ruby API for GPT, Claude, Gemini, and more.
 test_files: []

data/lib/ruby_llm/providers/bedrock/capabilities.rb

@@ -1,167 +0,0 @@
-# frozen_string_literal: true
-
-module RubyLLM
-  module Providers
-    class Bedrock
-      # Determines capabilities and pricing for AWS Bedrock models
-      module Capabilities
-        module_function
-
-        def context_window_for(model_id)
-          case model_id
-          when /anthropic\.claude-2/ then 100_000
-          else 200_000
-          end
-        end
-
-        def max_tokens_for(_model_id)
-          4_096
-        end
-
-        def input_price_for(model_id)
-          PRICES.dig(model_family(model_id), :input) || default_input_price
-        end
-
-        def output_price_for(model_id)
-          PRICES.dig(model_family(model_id), :output) || default_output_price
-        end
-
-        def supports_chat?(model_id)
-          model_id.match?(/anthropic\.claude/)
-        end
-
-        def supports_streaming?(model_id)
-          model_id.match?(/anthropic\.claude/)
-        end
-
-        def supports_images?(model_id)
-          model_id.match?(/anthropic\.claude/)
-        end
-
-        def supports_vision?(model_id)
-          model_id.match?(/anthropic\.claude/)
-        end
-
-        def supports_functions?(model_id)
-          model_id.match?(/anthropic\.claude/)
-        end
-
-        def supports_audio?(_model_id)
-          false
-        end
-
-        def supports_json_mode?(model_id)
-          model_id.match?(/anthropic\.claude/)
-        end
-
-        def format_display_name(model_id)
-          model_id.then { |id| humanize(id) }
-        end
-
-        def model_type(_model_id)
-          'chat'
-        end
-
-        def supports_structured_output?(_model_id)
-          false
-        end
-
-        # Model family patterns for capability lookup
-        MODEL_FAMILIES = {
-          /anthropic\.claude-3-opus/ => :claude3_opus,
-          /anthropic\.claude-3-sonnet/ => :claude3_sonnet,
-          /anthropic\.claude-3-5-sonnet/ => :claude3_sonnet,
-          /anthropic\.claude-3-7-sonnet/ => :claude3_sonnet,
-          /anthropic\.claude-3-haiku/ => :claude3_haiku,
-          /anthropic\.claude-3-5-haiku/ => :claude3_5_haiku,
-          /anthropic\.claude-v2/ => :claude2,
-          /anthropic\.claude-instant/ => :claude_instant
-        }.freeze
-
-        def model_family(model_id)
-          MODEL_FAMILIES.find { |pattern, _family| model_id.match?(pattern) }&.last || :other
-        end
-
-        # Pricing information for Bedrock models (per million tokens)
-        PRICES = {
-          claude3_opus: { input: 15.0, output: 75.0 },
-          claude3_sonnet: { input: 3.0, output: 15.0 },
-          claude3_haiku: { input: 0.25, output: 1.25 },
-          claude3_5_haiku: { input: 0.8, output: 4.0 },
-          claude2: { input: 8.0, output: 24.0 },
-          claude_instant: { input: 0.8, output: 2.4 }
-        }.freeze
-
-        def default_input_price
-          0.1
-        end
-
-        def default_output_price
-          0.2
-        end
-
-        def humanize(id)
-          id.tr('-', ' ')
-            .split('.')
-            .last
-            .split
-            .map(&:capitalize)
-            .join(' ')
-        end
-
-        def modalities_for(model_id)
-          modalities = {
-            input: ['text'],
-            output: ['text']
-          }
-
-          if model_id.match?(/anthropic\.claude/) && supports_vision?(model_id)
-            modalities[:input] << 'image'
-            modalities[:input] << 'pdf'
-          end
-
-          modalities
-        end
-
-        def capabilities_for(model_id)
-          capabilities = []
-
-          capabilities << 'streaming' if model_id.match?(/anthropic\.claude/)
-
-          capabilities << 'function_calling' if supports_functions?(model_id)
-
-          capabilities << 'reasoning' if model_id.match?(/claude-3-7/)
-
-          if model_id.match?(/claude-3\.5|claude-3-7/)
-            capabilities << 'batch'
-            capabilities << 'citations'
-          end
-
-          capabilities
-        end
-
-        def pricing_for(model_id)
-          family = model_family(model_id)
-          prices = PRICES.fetch(family, { input: default_input_price, output: default_output_price })
-
-          standard_pricing = {
-            input_per_million: prices[:input],
-            output_per_million: prices[:output]
-          }
-
-          batch_pricing = {
-            input_per_million: prices[:input] * 0.5,
-            output_per_million: prices[:output] * 0.5
-          }
-
-          {
-            text_tokens: {
-              standard: standard_pricing,
-              batch: batch_pricing
-            }
-          }
-        end
-      end
-    end
-  end
-end