docs: rewrite much of README.md for readability

stainless-app[bot] · stainless-app[bot] · commit f3c721e71de5 · 2025-05-14T21:53:43.000Z
diff --git a/README.md b/README.md
@@ -1,6 +1,6 @@
 # OpenAI Ruby API library
 
-The OpenAI Ruby library provides convenient access to the OpenAI REST API from any Ruby 3.2.0+ application.
+The OpenAI Ruby library provides convenient access to the OpenAI REST API from any Ruby 3.2.0+ application. It ships with comprehensive types & docstrings in Yard, RBS, and RBI – [see below](https://github.com/openai/openai-ruby#Sorbet) for usage with Sorbet. The standard library's `net/http` is used as the HTTP transport, with connection pooling via the `connection_pool` gem.
 
 ## Documentation
 
@@ -40,17 +40,19 @@ chat_completion = openai.chat.completions.create(
 puts(chat_completion)
 ```
 
-## Sorbet
-
-This library is written with [Sorbet type definitions](https://sorbet.org/docs/rbi). However, there is no runtime dependency on the `sorbet-runtime`.
+### Streaming
 
-When using sorbet, it is recommended to use model classes as below. This provides stronger type checking and tooling integration.
+We provide support for streaming responses using Server-Sent Events (SSE).
 
 ```ruby
-openai.chat.completions.create(
-  messages: [OpenAI::Chat::ChatCompletionUserMessageParam.new(role: "user", content: "Say this is a test")],
+stream = openai.chat.completions.stream_raw(
+  messages: [{role: "user", content: "Say this is a test"}],
   model: :"gpt-4.1"
 )
+
+stream.each do |completion|
+  puts(completion)
+end
 ```
 
 ### Pagination
@@ -72,49 +74,54 @@ page.auto_paging_each do |job|
 end
 ```
 
-### Streaming
-
-We provide support for streaming responses using Server-Sent Events (SSE).
+Alternatively, you can use the `#next_page?` and `#next_page` methods for more granular control working with pages.
 
 **coming soon:** `openai.chat.completions.stream` will soon come with Python SDK style higher level streaming responses support.
 
 ```ruby
-stream = openai.chat.completions.stream_raw(
-  messages: [{role: "user", content: "Say this is a test"}],
-  model: :"gpt-4.1"
-)
-
-stream.each do |completion|
-  print(completion.choices.first.delta.content)
+if page.next_page?
+  new_page = page.next_page
+  puts(new_page.data[0].id)
 end
 ```
 
 ### File uploads
 
-Request parameters that correspond to file uploads can be passed as `StringIO`, or a [`Pathname`](https://rubyapi.org/3.2/o/pathname) instance.
+Request parameters that correspond to file uploads can be passed as raw contents, a [`Pathname`](https://rubyapi.org/3.2/o/pathname) instance, [`StringIO`](https://rubyapi.org/3.2/o/stringio), or more.
 
 ```ruby
 require "pathname"
 
-# using `Pathname`, the file will be lazily read, without reading everything in to memory
+# Use `Pathname` to send the filename and/or avoid paging a large file into memory:
 file_object = openai.files.create(file: Pathname("input.jsonl"), purpose: "fine-tune")
 
-file = File.read("input.jsonl")
-# using `StringIO`, useful if you already have the data in memory
-file_object = openai.files.create(file: StringIO.new(file), purpose: "fine-tune")
+# Alternatively, pass file contents or a `StringIO` directly:
+file_object = openai.files.create(file: File.read("input.jsonl"), purpose: "fine-tune")
+
+# Or, to control the filename and/or content type:
+file = OpenAI::FilePart.new(File.read("input.jsonl"), filename: "input.jsonl", content_type: "…")
+file_object = openai.files.create(file: file, purpose: "fine-tune")
 
 puts(file_object.id)
 ```
 
-### Errors
+Note that you can also pass a raw `IO` descriptor, but this disables retries, as the library can't be sure if the descriptor is a file or pipe (which cannot be rewound).
+
+### Handling errors
 
 When the library is unable to connect to the API, or if the API returns a non-success status code (i.e., 4xx or 5xx response), a subclass of `OpenAI::Errors::APIError` will be thrown:
 
 ```ruby
 begin
   job = openai.fine_tuning.jobs.create(model: :"babbage-002", training_file: "file-abc123")
-rescue OpenAI::Errors::APIError => e
-  puts(e.status) # 400
+rescue OpenAI::Errors::APIConnectionError => e
+  puts("The server could not be reached")
+  puts(e.cause)  # an underlying Exception, likely raised within `net/http`
+rescue OpenAI::Errors::RateLimitError => e
+  puts("A 429 status code was received; we should back off a bit.")
+rescue OpenAI::Errors::APIStatusError => e
+  puts("Another non-200-range status code was received")
+  puts(e.status)
 end
 ```
 
@@ -158,11 +165,7 @@ openai.chat.completions.create(
 
 ### Timeouts
 
-By default, requests will time out after 600 seconds.
-
-Timeouts are applied separately to the initial connection and the overall request time, so in some cases a request could wait 2\*timeout seconds before it fails.
-
-You can use the `timeout` option to configure or disable this:
+By default, requests will time out after 600 seconds. You can use the timeout option to configure or disable this:
 
 ```ruby
 # Configure the default for all requests:
@@ -178,91 +181,133 @@ openai.chat.completions.create(
 )
 ```
 
-## Model DSL
+On timeout, `OpenAI::Errors::APITimeoutError` is raised.
 
-This library uses a simple DSL to represent request parameters and response shapes in `lib/openai/models`.
+Note that requests that time out are retried by default.
 
-With the right [editor plugins](https://shopify.github.io/ruby-lsp), you can ctrl-click on elements of the DSL to navigate around and explore the library.
+## Advanced concepts
 
-In all places where a `BaseModel` type is specified, vanilla Ruby `Hash` can also be used. For example, the following are interchangeable as arguments:
+### BaseModel
 
-```ruby
-# This has tooling readability, for auto-completion, static analysis, and goto definition with supported language services
-params = OpenAI::Models::Chat::CompletionCreateParams.new(
-  messages: [OpenAI::Chat::ChatCompletionUserMessageParam.new(role: "user", content: "Say this is a test")],
-  model: :"gpt-4.1"
-)
+All parameter and response objects inherit from `OpenAI::Internal::Type::BaseModel`, which provides several conveniences, including:
 
-# This also works
-params = {
-  messages: [{role: "user", content: "Say this is a test"}],
-  model: :"gpt-4.1"
-}
-```
+1. All fields, including unknown ones, are accessible with `obj[:prop]` syntax, and can be destructured with `obj => {prop: prop}` or pattern-matching syntax.
 
-## Editor support
+2. Structural equivalence for equality; if two API calls return the same values, comparing the responses with == will return true.
 
-A combination of [Shopify LSP](https://shopify.github.io/ruby-lsp) and [Solargraph](https://solargraph.org/) is recommended for non-[Sorbet](https://sorbet.org) users. The former is especially good at go to definition, while the latter has much better auto-completion support.
+3. Both instances and the classes themselves can be pretty-printed.
 
-## Advanced concepts
+4. Helpers such as `#to_h`, `#deep_to_h`, `#to_json`, and `#to_yaml`.
+
+### Making custom or undocumented requests
+
+#### Undocumented properties
 
-### Making custom/undocumented requests
+You can send undocumented parameters to any endpoint, and read undocumented response properties, like so:
+
+Note: the `extra_` parameters of the same name overrides the documented parameters.
+
+```ruby
+chat_completion =
+  openai.chat.completions.create(
+    messages: [{role: "user", content: "How can I get the name of the current day in JavaScript?"}],
+    model: :"gpt-4.1",
+    request_options: {
+      extra_query: {my_query_parameter: value},
+      extra_body: {my_body_parameter: value},
+      extra_headers: {"my-header": value}
+    }
+  )
+
+puts(chat_completion[:my_undocumented_property])
+```
 
 #### Undocumented request params
 
-If you want to explicitly send an extra param, you can do so with the `extra_query`, `extra_body`, and `extra_headers` under the `request_options:` parameter when making a requests as seen in examples above.
+If you want to explicitly send an extra param, you can do so with the `extra_query`, `extra_body`, and `extra_headers` under the `request_options:` parameter when making a request as seen in examples above.
 
 #### Undocumented endpoints
 
-To make requests to undocumented endpoints, you can make requests using `client.request`. Options on the client will be respected (such as retries) when making this request.
+To make requests to undocumented endpoints while retaining the benefit of auth, retries, and so on, you can make requests using `client.request`, like so:
 
 ```ruby
 response = client.request(
   method: :post,
   path: '/undocumented/endpoint',
   query: {"dog": "woof"},
   headers: {"useful-header": "interesting-value"},
-  body: {"he": "llo"},
+  body: {"hello": "world"}
 )
 ```
 
 ### Concurrency & connection pooling
 
-The `OpenAI::Client` instances are thread-safe, and should be re-used across multiple threads. By default, each `Client` have their own HTTP connection pool, with a maximum number of connections equal to thread count.
+The `OpenAI::Client` instances are threadsafe, but only are fork-safe when there are no in-flight HTTP requests.
 
-When the maximum number of connections has been checked out from the connection pool, the `Client` will wait for an in use connection to become available. The queue time for this mechanism is accounted for by the per-request timeout.
+Each instance of `OpenAI::Client` has its own HTTP connection pool with a default size of 99. As such, we recommend instantiating the client once per application in most settings.
 
-Unless otherwise specified, other classes in the SDK do not have locks protecting their underlying data structure.
+When all available connections from the pool are checked out, requests wait for a new connection to become available, with queue time counting towards the request timeout.
 
-Currently, `OpenAI::Client` instances are only fork-safe if there are no in-flight HTTP requests.
-
-### Sorbet
+Unless otherwise specified, other classes in the SDK do not have locks protecting their underlying data structure.
 
-#### Enums
+## Sorbet
 
-Sorbet's typed enums require sub-classing of the [`T::Enum` class](https://sorbet.org/docs/tenum) from the `sorbet-runtime` gem.
+This library provides comprehensive [RBI](https://sorbet.org/docs/rbi) definitions, and has no dependency on sorbet-runtime.
 
-Since this library does not depend on `sorbet-runtime`, it uses a [`T.all` intersection type](https://sorbet.org/docs/intersection-types) with a ruby primitive type to construct a "tagged alias" instead.
+You can provide typesafe request parameters like so:
 
 ```ruby
-module OpenAI::ChatModel
-  # This alias aids language service driven navigation.
-  TaggedSymbol = T.type_alias { T.all(Symbol, OpenAI::ChatModel) }
-end
+openai.chat.completions.create(
+  messages: [OpenAI::Chat::ChatCompletionUserMessageParam.new(role: "user", content: "Say this is a test")],
+  model: :"gpt-4.1"
+)
 ```
 
-#### Argument passing trick
-
-It is possible to pass a compatible model / parameter class to a method that expects keyword arguments by using the `**` splat operator.
+Or, equivalently:
 
 ```ruby
-params = OpenAI::Models::Chat::CompletionCreateParams.new(
+# Hashes work, but are not typesafe:
+openai.chat.completions.create(
+  messages: [{role: "user", content: "Say this is a test"}],
+  model: :"gpt-4.1"
+)
+
+# You can also splat a full Params class:
+params = OpenAI::Chat::CompletionCreateParams.new(
   messages: [OpenAI::Chat::ChatCompletionUserMessageParam.new(role: "user", content: "Say this is a test")],
   model: :"gpt-4.1"
 )
 openai.chat.completions.create(**params)
 ```
 
+### Enums
+
+Since this library does not depend on `sorbet-runtime`, it cannot provide [`T::Enum`](https://sorbet.org/docs/tenum) instances. Instead, we provide "tagged symbols" instead, which is always a primitive at runtime:
+
+```ruby
+# :low
+puts(OpenAI::ReasoningEffort::LOW)
+
+# Revealed type: `T.all(OpenAI::ReasoningEffort, Symbol)`
+T.reveal_type(OpenAI::ReasoningEffort::LOW)
+```
+
+Enum parameters have a "relaxed" type, so you can either pass in enum constants or their literal value:
+
+```ruby
+# Using the enum constants preserves the tagged type information:
+openai.chat.completions.create(
+  reasoning_effort: OpenAI::ReasoningEffort::LOW,
+  # …
+)
+
+# Literal values is also permissible:
+openai.chat.completions.create(
+  reasoning_effort: :low,
+  # …
+)
+```
+
 ## Versioning
 
 This package follows [SemVer](https://semver.org/spec/v2.0.0.html) conventions. As the library is in initial development and has a major version of `0`, APIs may change at any time.
diff --git a/lib/openai/internal/type/base_model.rb b/lib/openai/internal/type/base_model.rb
@@ -386,6 +386,14 @@ def deep_to_h = self.class.recursively_to_h(@data, convert: false)
         # @param keys [Array<Symbol>, nil]
         #
         # @return [Hash{Symbol=>Object}]
+        #
+        # @example
+        #   # `comparison_filter` is a `OpenAI::ComparisonFilter`
+        #   comparison_filter => {
+        #     key: key,
+        #     type: type,
+        #     value: value
+        #   }
         def deconstruct_keys(keys)
           (keys || self.class.known_fields.keys)
             .filter_map do |k|
