ruby_llm 1.1.2 → 1.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d3226fdaf84de31b1deefae4aa85ca721d4fc6ce2a84572d211a57a0acbf1cb3
-  data.tar.gz: 8846beaefacc53539987eaf6b619e4833267a928f688ad6afc4277180eaac7b2
+  metadata.gz: 2f06ce431337dc189e6172b0c98ed897fdba930200f3f118c39c15f4527ec135
+  data.tar.gz: 18f8ff36e7ee18cbee315e66db4b8f04619c98595a5de3b73d215bed248ca0d4
 SHA512:
-  metadata.gz: 1556e816f7ffdb9323b53067f6faefc9c96eb0e116740b690dfa96afdc39c96d0b294efe4b8a363279bf6d43c907d4e6ebdf92cde0dc996218961627a655a626
-  data.tar.gz: 75def96179be6d1d44cce576a698aaeb7bf5cf11ac60d0699f12de7d7dfa32f9a71906821ec43d1f77b98fd152156af41d10b89e832406601f42239872950a8c
+  metadata.gz: 42f7603cfec24fa6cc59b1186d2d6a90af9e9076eb79124ac5ce09d73000fbcdb931ab90cafe21bf95b39417a52ff000ca9d02ba51c8a78877d1a1f47b70866f
+  data.tar.gz: 8513b6774ef3d745e7bbc8947f856608d13ade163ca658139c2992e923ad08d5ee00b99275cbfe762591182897b73db7bfc879fae8db7b8ce2f3ba1fea5ee235

data/README.md

@@ -118,7 +118,7 @@ end
 ## Have great conversations
 
 ```ruby
-# Start a chat with the default model (GPT-4o-mini)
+# Start a chat with the default model (gpt-4.1-nano)
 chat = RubyLLM.chat
 
 # Or specify what you want
@@ -169,7 +169,7 @@ class ToolCall < ApplicationRecord
 end
 
 # In a background job
-chat = Chat.create! model_id: "gpt-4o-mini"
+chat = Chat.create! model_id: "gpt-4.1-nano"
 
 # Set personality or behavior with instructions (aka system prompts) - they're persisted too!
 chat.with_instructions "You are a friendly Ruby expert who loves to help beginners"

data/lib/ruby_llm/chat.rb

@@ -8,14 +8,18 @@ module RubyLLM
   #   chat = RubyLLM.chat
   #   chat.ask "What's the best way to learn Ruby?"
   #   chat.ask "Can you elaborate on that?"
-  class Chat
+  class Chat # rubocop:disable Metrics/ClassLength
     include Enumerable
 
     attr_reader :model, :messages, :tools
 
-    def initialize(model: nil, provider: nil)
+    def initialize(model: nil, provider: nil, assume_model_exists: false) # rubocop:disable Metrics/MethodLength
+      if assume_model_exists && !provider
+        raise ArgumentError, 'Provider must be specified if assume_model_exists is true'
+      end
+
       model_id = model || RubyLLM.config.default_model
-      with_model(model_id, provider: provider)
+      with_model(model_id, provider: provider, assume_exists: assume_model_exists)
       @temperature = 0.7
       @messages = []
       @tools = {}
@@ -54,9 +58,18 @@ module RubyLLM
       self
     end
 
-    def with_model(model_id, provider: nil)
-      @model = Models.find model_id, provider
-      @provider = Provider.providers[@model.provider.to_sym] || raise(Error, "Unknown provider: #{@model.provider}")
+    def with_model(model_id, provider: nil, assume_exists: false) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength
+      if assume_exists
+        raise ArgumentError, 'Provider must be specified if assume_exists is true' unless provider
+
+        @provider = Provider.providers[provider.to_sym] || raise(Error, "Unknown provider: #{provider.to_sym}")
+        @model = Struct.new(:id, :provider, :supports_functions, :supports_vision).new(model_id, provider, true, true)
+        RubyLLM.logger.warn "Assuming model '#{model_id}' exists for provider '#{provider}'. " \
+                            'Capabilities may not be accurately reflected.'
+      else
+        @model = Models.find model_id, provider
+        @provider = Provider.providers[@model.provider.to_sym] || raise(Error, "Unknown provider: #{@model.provider}")
+      end
       self
     end
 

data/lib/ruby_llm/configuration.rb

@@ -12,6 +12,7 @@ module RubyLLM
   class Configuration
     # Provider-specific configuration
     attr_accessor :openai_api_key,
+                  :openai_api_base,
                   :anthropic_api_key,
                   :gemini_api_key,
                   :deepseek_api_key,
@@ -39,7 +40,7 @@ module RubyLLM
       @retry_interval_randomness = 0.5
 
       # Default models
-      @default_model = 'gpt-4o-mini'
+      @default_model = 'gpt-4.1-nano'
       @default_embedding_model = 'text-embedding-3-small'
       @default_image_model = 'dall-e-3'
     end

data/lib/ruby_llm/models.json

@@ -270,36 +270,6 @@
       ]
     }
   },
-  {
-    "id": "anthropic.claude-3-7-sonnet-20250219-v1:0",
-    "created_at": null,
-    "display_name": "Claude 3.7 Sonnet",
-    "provider": "bedrock",
-    "context_window": 200000,
-    "max_tokens": 4096,
-    "type": "chat",
-    "family": "claude3_sonnet",
-    "supports_vision": true,
-    "supports_functions": true,
-    "supports_json_mode": true,
-    "input_price_per_million": 3.0,
-    "output_price_per_million": 15.0,
-    "metadata": {
-      "provider_name": "Anthropic",
-      "customizations_supported": [],
-      "inference_configurations": [
-        "INFERENCE_PROFILE"
-      ],
-      "response_streaming_supported": true,
-      "input_modalities": [
-        "TEXT",
-        "IMAGE"
-      ],
-      "output_modalities": [
-        "TEXT"
-      ]
-    }
-  },
   {
     "id": "anthropic.claude-3-haiku-20240307-v1:0",
     "created_at": null,
@@ -1918,7 +1888,8 @@
       "output_token_limit": 65536,
       "supported_generation_methods": [
         "generateContent",
-        "countTokens"
+        "countTokens",
+        "createCachedContent"
       ]
     }
   },
@@ -1943,7 +1914,8 @@
       "output_token_limit": 65536,
       "supported_generation_methods": [
         "generateContent",
-        "countTokens"
+        "countTokens",
+        "createCachedContent"
       ]
     }
   },
@@ -1968,7 +1940,8 @@
       "output_token_limit": 65536,
       "supported_generation_methods": [
         "generateContent",
-        "countTokens"
+        "countTokens",
+        "createCachedContent"
       ]
     }
   },
@@ -1993,7 +1966,8 @@
       "output_token_limit": 65536,
       "supported_generation_methods": [
         "generateContent",
-        "countTokens"
+        "countTokens",
+        "createCachedContent"
       ]
     }
   },
@@ -2068,7 +2042,8 @@
       "output_token_limit": 65536,
       "supported_generation_methods": [
         "generateContent",
-        "countTokens"
+        "countTokens",
+        "createCachedContent"
       ]
     }
   },
@@ -3082,6 +3057,31 @@
       ]
     }
   },
+  {
+    "id": "learnlm-2.0-flash-experimental",
+    "created_at": null,
+    "display_name": "LearnLM 2.0 Flash Experimental",
+    "provider": "gemini",
+    "context_window": 1048576,
+    "max_tokens": 32768,
+    "type": "chat",
+    "family": "other",
+    "supports_vision": true,
+    "supports_functions": true,
+    "supports_json_mode": true,
+    "input_price_per_million": 0.075,
+    "output_price_per_million": 0.3,
+    "metadata": {
+      "version": "2.0",
+      "description": "LearnLM 2.0 Flash Experimental",
+      "input_token_limit": 1048576,
+      "output_token_limit": 32768,
+      "supported_generation_methods": [
+        "generateContent",
+        "countTokens"
+      ]
+    }
+  },
   {
     "id": "o1",
     "created_at": "2024-12-16T20:03:36+01:00",
@@ -3272,6 +3272,44 @@
       "owned_by": "system"
     }
   },
+  {
+    "id": "o4-mini",
+    "created_at": "2025-04-09T21:02:31+02:00",
+    "display_name": "O4 Mini",
+    "provider": "openai",
+    "context_window": 4096,
+    "max_tokens": 16384,
+    "type": "chat",
+    "family": "other",
+    "supports_vision": false,
+    "supports_functions": false,
+    "supports_json_mode": false,
+    "input_price_per_million": 0.5,
+    "output_price_per_million": 1.5,
+    "metadata": {
+      "object": "model",
+      "owned_by": "system"
+    }
+  },
+  {
+    "id": "o4-mini-2025-04-16",
+    "created_at": "2025-04-08T19:31:46+02:00",
+    "display_name": "O4 Mini 20250416",
+    "provider": "openai",
+    "context_window": 4096,
+    "max_tokens": 16384,
+    "type": "chat",
+    "family": "other",
+    "supports_vision": false,
+    "supports_functions": false,
+    "supports_json_mode": false,
+    "input_price_per_million": 0.5,
+    "output_price_per_million": 1.5,
+    "metadata": {
+      "object": "model",
+      "owned_by": "system"
+    }
+  },
   {
     "id": "omni-moderation-2024-09-26",
     "created_at": "2024-11-27T20:07:46+01:00",
@@ -3493,6 +3531,36 @@
       "owned_by": "system"
     }
   },
+  {
+    "id": "us.anthropic.claude-3-7-sonnet-20250219-v1:0",
+    "created_at": null,
+    "display_name": "Claude 3.7 Sonnet",
+    "provider": "bedrock",
+    "context_window": 200000,
+    "max_tokens": 4096,
+    "type": "chat",
+    "family": "claude3_sonnet",
+    "supports_vision": true,
+    "supports_functions": true,
+    "supports_json_mode": true,
+    "input_price_per_million": 3.0,
+    "output_price_per_million": 15.0,
+    "metadata": {
+      "provider_name": "Anthropic",
+      "customizations_supported": [],
+      "inference_configurations": [
+        "INFERENCE_PROFILE"
+      ],
+      "response_streaming_supported": true,
+      "input_modalities": [
+        "TEXT",
+        "IMAGE"
+      ],
+      "output_modalities": [
+        "TEXT"
+      ]
+    }
+  },
   {
     "id": "veo-2.0-generate-001",
     "created_at": null,

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

@@ -42,13 +42,20 @@ module RubyLLM
 
         def base_model_attributes(model_id, model, slug)
           {
-            id: model_id,
+            id: model_id_with_prefix(model_id, model),
             created_at: nil,
             display_name: model['modelName'] || capabilities.format_display_name(model_id),
             provider: slug
           }
         end
 
+        def model_id_with_prefix(model_id, model)
+          return model_id unless model['inferenceTypesSupported']&.include?('INFERENCE_PROFILE')
+          return model_id if model['inferenceTypesSupported']&.include?('ON_DEMAND')
+
+          "us.#{model_id}"
+        end
+
         def capability_attributes(model_id, capabilities)
           {
             context_window: capabilities.context_window_for(model_id),

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

@@ -215,7 +215,7 @@ module RubyLLM
         end
 
         def normalize_temperature(temperature, model_id)
-          if model_id.match?(/^o[13]/)
+          if model_id.match?(/^o\d/)
             RubyLLM.logger.debug "Model #{model_id} requires temperature=1.0, ignoring provided value"
             1.0
           else

data/lib/ruby_llm/providers/openai.rb

@@ -29,7 +29,7 @@ module RubyLLM
       module_function
 
       def api_base
-        'https://api.openai.com/v1'
+        RubyLLM.config.openai_api_base || 'https://api.openai.com/v1'
       end
 
       def headers

data/lib/ruby_llm/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module RubyLLM
-  VERSION = '1.1.2'
+  VERSION = '1.2.0'
 end

data/lib/ruby_llm.rb

@@ -30,8 +30,8 @@ module RubyLLM
   class Error < StandardError; end
 
   class << self
-    def chat(model: nil, provider: nil)
-      Chat.new(model: model, provider: provider)
+    def chat(model: nil, provider: nil, assume_model_exists: false)
+      Chat.new(model:, provider:, assume_model_exists:)
     end
 
     def embed(...)

data/lib/tasks/models_docs.rake

@@ -86,10 +86,18 @@ namespace :models do # rubocop:disable Metrics/BlockLength
       ---
 
       # Available Models
+      {: .no_toc }
 
       This guide lists all models available in RubyLLM, automatically generated from the current model registry.
+      {: .fs-6 .fw-300 }
 
-      _Last updated: #{Time.now.utc.strftime('%Y-%m-%d')}_
+      ## Table of contents
+      {: .no_toc .text-delta }
+
+      1. TOC
+      {:toc}
+
+      ---
 
       ## Contributing
 
@@ -115,6 +123,10 @@ namespace :models do # rubocop:disable Metrics/BlockLength
       For more information about working with models, see the [Working with Models](/guides/models) guide.
 
       ## Models by Type
+      {: .d-inline-block }
+
+      Last updated: #{Time.now.utc.strftime('%Y-%m-%d')}
+      {: .label .label-green }
 
       ### Chat Models (#{RubyLLM.models.chat_models.count})
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ruby_llm
 version: !ruby/object:Gem::Version
-  version: 1.1.2
+  version: 1.2.0
 platform: ruby
 authors:
 - Carmine Paolino
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2025-04-15 00:00:00.000000000 Z
+date: 2025-04-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: base64