ruby_llm 0.1.0.pre40 → 0.1.0.pre42

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: db95ae0d7103334e7760659d58200e1cc882b970a2d3f67f56180a9879bad976
-  data.tar.gz: f2a95fb00c476cbea2bf060ff5502c4fb61ed2d652578133cf7eee6ff4c7297f
+  metadata.gz: 2dac5b1b15a5d840d74ac63af791044a26669e0d41b0512d37236530cdd89693
+  data.tar.gz: b737c0061790066680424051afd3b9ea75d9d63f014d1684261dde81c1bedf37
 SHA512:
-  metadata.gz: 3a3624f1c5d8184d5f36aca807528b4bc627a66aebe3829ff9e90bb62348a12863b764352d4b466931c0ef3469c1d952832a1b7726f93231b72324b13228ffb1
-  data.tar.gz: 4bef0ae0373be0c2cd8d4702ddfc2f7a76883fe7c60e5ae5ef501c6f808744328919fba3f16dedf075d6f1cb1585120f46103415579a046306fd1a4d1c81f20b
+  metadata.gz: a90362b3d2dbe5b9f5c54851d94646273b343a32eef452528903026ea2e1add335a105273b0efc17a5ebde1da15421d894029ebe401e94c511531a1e8d2f706b
+  data.tar.gz: 504e5dd4e3692db83b6b9589e26e33f553865234b59264d05aec1f495e403e8e28eb2b7dfc2b2258047df07dd04f3740005b44dee8a874927b12a55f045fae6e

data/.rspec_status

@@ -7,6 +7,8 @@ example_id                                         | status | run_time        |
 ./spec/ruby_llm/chat_content_spec.rb[1:1:3]        | passed | 2.54 seconds    |
 ./spec/ruby_llm/chat_content_spec.rb[1:2:1]        | passed | 2.77 seconds    |
 ./spec/ruby_llm/chat_content_spec.rb[1:2:2]        | passed | 2.1 seconds     |
+./spec/ruby_llm/chat_pdf_spec.rb[1:1:1]            | passed | 7.75 seconds    |
+./spec/ruby_llm/chat_pdf_spec.rb[1:1:2]            | passed | 13.88 seconds   |
 ./spec/ruby_llm/chat_spec.rb[1:1:1:1]              | passed | 1.02 seconds    |
 ./spec/ruby_llm/chat_spec.rb[1:1:1:2]              | passed | 3.95 seconds    |
 ./spec/ruby_llm/chat_spec.rb[1:1:2:1]              | passed | 0.4854 seconds  |
@@ -32,7 +34,8 @@ example_id                                         | status | run_time        |
 ./spec/ruby_llm/embeddings_spec.rb[1:1:1:2]        | passed | 0.43632 seconds |
 ./spec/ruby_llm/embeddings_spec.rb[1:1:2:1]        | passed | 0.65614 seconds |
 ./spec/ruby_llm/embeddings_spec.rb[1:1:2:2]        | passed | 2.16 seconds    |
-./spec/ruby_llm/error_handling_spec.rb[1:1]        | passed | 0.19586 seconds |
-./spec/ruby_llm/image_generation_spec.rb[1:1:1]    | passed | 12.77 seconds   |
-./spec/ruby_llm/image_generation_spec.rb[1:1:2]    | passed | 18.13 seconds   |
-./spec/ruby_llm/image_generation_spec.rb[1:1:3]    | passed | 0.00035 seconds |
+./spec/ruby_llm/error_handling_spec.rb[1:1]        | passed | 0.29366 seconds |
+./spec/ruby_llm/image_generation_spec.rb[1:1:1]    | passed | 11.61 seconds   |
+./spec/ruby_llm/image_generation_spec.rb[1:1:2]    | passed | 17.63 seconds   |
+./spec/ruby_llm/image_generation_spec.rb[1:1:3]    | passed | 8.77 seconds    |
+./spec/ruby_llm/image_generation_spec.rb[1:1:4]    | passed | 0.00319 seconds |

data/README.md

@@ -16,7 +16,6 @@ A delightful Ruby way to work with AI. Chat in text, analyze and generate images
   <a href="https://badge.fury.io/rb/ruby_llm"><img src="https://badge.fury.io/rb/ruby_llm.svg" alt="Gem Version" /></a>
   <a href="https://github.com/testdouble/standard"><img src="https://img.shields.io/badge/code_style-standard-brightgreen.svg" alt="Ruby Style Guide" /></a>
   <a href="https://rubygems.org/gems/ruby_llm"><img alt="Gem Total Downloads" src="https://img.shields.io/gem/dt/ruby_llm"></a>
-  <a href="https://github.com/crmne/ruby_llm/actions/workflows/cicd.yml"><img src="https://github.com/crmne/ruby_llm/actions/workflows/cicd.yml/badge.svg" alt="CI" /></a>
   <a href="https://codecov.io/gh/crmne/ruby_llm"><img src="https://codecov.io/gh/crmne/ruby_llm/branch/main/graph/badge.svg" alt="codecov" /></a>
 </p>
 
@@ -28,6 +27,7 @@ A delightful Ruby way to work with AI. Chat in text, analyze and generate images
 - 🎵 **Audio Analysis** - Get audio transcription and understanding with `chat.ask "what's said here?", with: { audio: "clip.wav" }`
 - 👁️ **Vision Understanding** - Let AIs analyze images with a simple `chat.ask "what's in this?", with: { image: "photo.jpg" }`
 - 🌊 **Streaming** - Real-time responses with proper Ruby streaming with `chat.ask "hello" do |chunk| puts chunk.content end`
+- 📄 **PDF Analysis** - Analyze PDF documents directly with `chat.ask "What's in this?", with: { pdf: "document.pdf" }`
 - 🚂 **Rails Integration** - Persist chats and messages with ActiveRecord with `acts_as_{chat|message|tool_call}`
 - 🛠️ **Tool Support** - Give AIs access to your Ruby code with `chat.with_tool(Calculator).ask "what's 2+2?"`
 - 🎨 **Paint with AI** - Create images as easily as `RubyLLM.paint "a sunset over mountains"`
@@ -116,6 +116,14 @@ chat.ask "What's being said in this recording?", with: { audio: "meeting.wav" }
 # Combine multiple pieces of content
 chat.ask "Compare these diagrams", with: { image: ["diagram1.png", "diagram2.png"] }
 
+# Ask about PDFs
+
+chat = RubyLLM.chat(model: 'claude-3-7-sonnet-20250219')
+chat.ask "Summarize this research paper", with: { pdf: "research.pdf" }
+
+# Multiple PDFs work too
+chat.ask "Compare these contracts", with: { pdf: ["contract1.pdf", "contract2.pdf"] }
+
 # Check token usage
 last_message = chat.messages.last
 puts "Conversation used #{last_message.input_tokens} input tokens and #{last_message.output_tokens} output tokens"
@@ -440,6 +448,21 @@ pp chat.messages.map(&:role)
 #=> [:user, :assistant, :tool, :assistant]
 ```
 
+## Provider Comparison
+
+| Feature | OpenAI | Anthropic | Google | DeepSeek |
+|---------|--------|-----------|--------|----------|
+| Chat | ✅ GPT-4o, GPT-3.5 | ✅ Claude 3.7, 3.5, 3 | ✅ Gemini 2.0, 1.5 | ✅ DeepSeek Chat, Reasoner |
+| Vision | ✅ GPT-4o, GPT-4 | ✅ All Claude 3 models | ✅ Gemini 2.0, 1.5 | ❌ |
+| Audio | ✅ GPT-4o-audio, Whisper | ❌ | ✅ Gemini models | ❌ |
+| PDF Analysis | ❌ | ✅ All Claude 3 models | ✅ Gemini models | ❌ |
+| Function Calling | ✅ Most models | ✅ Claude 3 models | ✅ Gemini models (except Lite) | ✅ |
+| JSON Mode | ✅ Most recent models | ✅ Claude 3 models | ✅ Gemini models | ❌ |
+| Image Generation | ✅ DALL-E 3 | ❌ | ✅ Imagen | ❌ |
+| Embeddings | ✅ text-embedding-3 | ❌ | ✅ text-embedding-004 | ❌ |
+| Context Size | ⭐ Up to 200K (o1) | ⭐ 200K tokens | ⭐ Up to 2M tokens | 64K tokens |
+| Streaming | ✅ | ✅ | ✅ | ✅ |
+
 ## Development
 
 After checking out the repo, run `bin/setup` to install dependencies. Then, run `bin/console` for an interactive prompt.

data/lib/ruby_llm/content.rb

@@ -5,7 +5,7 @@ module RubyLLM
   # Stores data in a standard internal format, letting providers
   # handle their own formatting needs.
   class Content
-    def initialize(text = nil, attachments = {})
+    def initialize(text = nil, attachments = {}) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength
       @parts = []
       @parts << { type: 'text', text: text } unless text.nil? || text.empty?
 
@@ -16,6 +16,10 @@ module RubyLLM
       Array(attachments[:audio]).each do |source|
         @parts << attach_audio(source)
       end
+
+      Array(attachments[:pdf]).each do |source|
+        @parts << attach_pdf(source)
+      end
     end
 
     def to_a
@@ -64,6 +68,25 @@ module RubyLLM
       }
     end
 
+    def attach_pdf(source)
+      source = File.expand_path(source) unless source.start_with?('http')
+
+      pdf_data = {
+        type: 'pdf',
+        source: source
+      }
+
+      # For local files, validate they exist
+      unless source.start_with?('http')
+        raise Error, "PDF file not found: #{source}" unless File.exist?(source)
+
+        # Preload file content for providers that need it
+        pdf_data[:content] = File.read(source)
+      end
+
+      pdf_data
+    end
+
     def encode_file(source)
       if source.start_with?('http')
         response = Faraday.get(source)