anthropic 1.80.0 → 2.3.0

data/rbi/anthropic/resources/beta/messages.rbi

@@ -10,7 +10,13 @@ module Anthropic
         sig do
           params(
             max_tokens: Integer,
-            messages: T::Array[Anthropic::Beta::BetaMessageParam::OrHash],
+            messages:
+              T::Array[
+                T.any(
+                  Anthropic::Beta::BetaMessageParam::Variants,
+                  Anthropic::Internal::AnyHash
+                )
+              ],
             model: T.any(Anthropic::Model::OrSymbol, String),
             container:
               T.nilable(
@@ -27,7 +33,6 @@ module Anthropic
               Anthropic::Beta::MessageCreateParams::ServiceTier::OrSymbol,
             stop_sequences: T::Array[String],
             system_: Anthropic::Beta::MessageCreateParams::System::Variants,
-            temperature: Float,
             thinking:
               T.any(
                 Anthropic::Beta::BetaThinkingConfigEnabled::OrHash,
@@ -59,9 +64,9 @@ module Anthropic
                   Anthropic::Beta::BetaWebFetchTool20250910::OrHash
                 )
               ],
-            top_k: Integer,
-            top_p: Float,
             betas: T::Array[T.any(String, Anthropic::AnthropicBeta::OrSymbol)],
+            max_iterations: T.nilable(Integer),
+            thread: T.nilable(FalseClass),
             stream: T.noreturn,
             request_options: Anthropic::RequestOptions::OrHash
           ).returns(Anthropic::Helpers::Tools::Runner)
@@ -136,8 +141,7 @@ module Anthropic
           #
           # Note that if you want to include a
           # [system prompt](https://docs.claude.com/en/docs/system-prompts), you can use the
-          # top-level `system` parameter — there is no `"system"` role for input messages in
-          # the Messages API.
+          # top-level `system` parameter.
           #
           # There is a limit of 100,000 messages in a single request.
           messages:,
@@ -178,15 +182,6 @@ module Anthropic
           # as specifying a particular goal or role. See our
           # [guide to system prompts](https://docs.claude.com/en/docs/system-prompts).
           system_: nil,
-          # Body param: Amount of randomness injected into the response.
-          #
-          # Defaults to `1.0`. Ranges from `0.0` to `1.0`. Use `temperature` closer to `0.0`
-          # for analytical / multiple choice, and closer to `1.0` for creative and
-          # generative tasks.
-          #
-          # Note that even with `temperature` of `0.0`, the results will not be fully
-          # deterministic.
-          temperature: nil,
           # Body param: Configuration for enabling Claude's extended thinking.
           #
           # When enabled, responses include `thinking` content blocks showing Claude's
@@ -276,26 +271,15 @@ module Anthropic
           #
           # See our [guide](https://docs.claude.com/en/docs/tool-use) for more details.
           tools: nil,
-          # Body param: Only sample from the top K options for each subsequent token.
-          #
-          # Used to remove "long tail" low probability responses.
-          # [Learn more technical details here](https://towardsdatascience.com/how-to-sample-from-language-models-682bceb97277).
-          #
-          # Recommended for advanced use cases only. You usually only need to use
-          # `temperature`.
-          top_k: nil,
-          # Body param: Use nucleus sampling.
-          #
-          # In nucleus sampling, we compute the cumulative distribution over all the options
-          # for each subsequent token in decreasing probability order and cut it off once it
-          # reaches a particular probability specified by `top_p`. You should either alter
-          # `temperature` or `top_p`, but not both.
-          #
-          # Recommended for advanced use cases only. You usually only need to use
-          # `temperature`.
-          top_p: nil,
           # Header param: Optional header to specify the beta version(s) you want to use.
           betas: nil,
+          # The most requests the runner sends before it stops. No limit when `nil`.
+          max_iterations: nil,
+          # The runner sends its requests on a message thread, so that each request after
+          # the first carries only the messages that are new. `false` sends the whole
+          # history with every request instead. `nil` leaves the thread on, and anything
+          # else raises `ArgumentError`.
+          thread: nil,
           # There is no need to provide `stream:`. Instead, use `#stream_raw` or `#create`
           # for streaming and non-streaming use cases, respectively.
           stream: false,
@@ -303,6 +287,152 @@ module Anthropic
         )
         end
 
+        # Creates a message thread: a conversation that the API keeps between requests, so
+        # that each turn sends only the messages that are new. It sends no request.
+        #
+        # Takes the keywords of `#create` except `thread` and `stream`.
+        sig do
+          params(
+            max_tokens: Integer,
+            model: T.any(Anthropic::Model::OrSymbol, String),
+            messages:
+              T::Array[
+                T.any(
+                  Anthropic::Beta::BetaUserMessageParam::OrHash,
+                  Anthropic::Beta::BetaAssistantMessageParam::OrHash,
+                  Anthropic::Beta::BetaSystemMessageParam::OrHash
+                )
+              ],
+            cache_control:
+              T.nilable(Anthropic::Beta::BetaCacheControlEphemeral::OrHash),
+            compaction:
+              T.nilable(Anthropic::Beta::BetaCompactionConfig::OrHash),
+            container:
+              T.nilable(
+                T.any(Anthropic::Beta::BetaContainerParams::OrHash, String)
+              ),
+            context_management:
+              T.nilable(Anthropic::Beta::BetaContextManagementConfig::OrHash),
+            diagnostics:
+              T.nilable(Anthropic::Beta::BetaDiagnosticsParam::OrHash),
+            fallback_credit_token:
+              T.nilable(
+                T.any(
+                  Anthropic::Beta::BetaFallbackCreditTokenParam::OrHash,
+                  String
+                )
+              ),
+            fallbacks:
+              T.nilable(
+                T.any(
+                  Symbol,
+                  T::Array[Anthropic::Beta::BetaFallbackParam::OrHash]
+                )
+              ),
+            inference_geo: T.nilable(String),
+            mcp_servers:
+              T::Array[
+                Anthropic::Beta::BetaRequestMCPServerURLDefinition::OrHash
+              ],
+            metadata: Anthropic::Beta::BetaMetadata::OrHash,
+            output_config: Anthropic::Beta::BetaOutputConfig::OrHash,
+            output_format:
+              T.nilable(Anthropic::Beta::BetaJSONOutputFormat::OrHash),
+            service_tier:
+              Anthropic::Beta::MessageCreateParams::ServiceTier::OrSymbol,
+            speed:
+              T.nilable(Anthropic::Beta::MessageCreateParams::Speed::OrSymbol),
+            stop_sequences: T::Array[String],
+            system_: Anthropic::Beta::MessageCreateParams::System::Variants,
+            thinking:
+              T.any(
+                Anthropic::Beta::BetaThinkingConfigEnabled::OrHash,
+                Anthropic::Beta::BetaThinkingConfigDisabled::OrHash,
+                Anthropic::Beta::BetaThinkingConfigBetweenTools::OrHash,
+                Anthropic::Beta::BetaThinkingConfigAdaptive::OrHash
+              ),
+            tool_choice:
+              T.any(
+                Anthropic::Beta::BetaToolChoiceAuto::OrHash,
+                Anthropic::Beta::BetaToolChoiceAny::OrHash,
+                Anthropic::Beta::BetaToolChoiceTool::OrHash,
+                Anthropic::Beta::BetaToolChoiceNone::OrHash
+              ),
+            tools:
+              T::Array[
+                T.any(
+                  Anthropic::Beta::BetaTool::OrHash,
+                  Anthropic::Beta::BetaToolBash20241022::OrHash,
+                  Anthropic::Beta::BetaToolBash20250124::OrHash,
+                  Anthropic::Beta::BetaCodeExecutionTool20250522::OrHash,
+                  Anthropic::Beta::BetaCodeExecutionTool20250825::OrHash,
+                  Anthropic::Beta::BetaCodeExecutionTool20260120::OrHash,
+                  Anthropic::Beta::BetaCodeExecutionTool20260521::OrHash,
+                  Anthropic::Beta::BetaBrowserToolset20260801::OrHash,
+                  Anthropic::Beta::BetaToolComputerUse20241022::OrHash,
+                  Anthropic::Beta::BetaMemoryTool20250818::OrHash,
+                  Anthropic::Beta::BetaToolComputerUse20250124::OrHash,
+                  Anthropic::Beta::BetaToolTextEditor20241022::OrHash,
+                  Anthropic::Beta::BetaToolComputerUse20251124::OrHash,
+                  Anthropic::Beta::BetaComputerToolset20260801::OrHash,
+                  Anthropic::Beta::BetaToolTextEditor20250124::OrHash,
+                  Anthropic::Beta::BetaToolTextEditor20250429::OrHash,
+                  Anthropic::Beta::BetaToolTextEditor20250728::OrHash,
+                  Anthropic::Beta::BetaWebSearchTool20250305::OrHash,
+                  Anthropic::Beta::BetaWebFetchTool20250910::OrHash,
+                  Anthropic::Beta::BetaWebSearchTool20260209::OrHash,
+                  Anthropic::Beta::BetaWebFetchTool20260209::OrHash,
+                  Anthropic::Beta::BetaWebFetchTool20260309::OrHash,
+                  Anthropic::Beta::BetaWebSearchTool20260318::OrHash,
+                  Anthropic::Beta::BetaWebFetchTool20260318::OrHash,
+                  Anthropic::Beta::BetaAdvisorTool20260301::OrHash,
+                  Anthropic::Beta::BetaToolSearchToolBm25_20251119::OrHash,
+                  Anthropic::Beta::BetaToolSearchToolRegex20251119::OrHash,
+                  Anthropic::Beta::BetaMCPToolset::OrHash
+                )
+              ],
+            betas: T::Array[T.any(Anthropic::AnthropicBeta::OrSymbol, String)],
+            user_profile_id: String,
+            workspace_id: String,
+            unsupported_request_behavior: Symbol,
+            request_options: Anthropic::RequestOptions::OrHash
+          ).returns(Anthropic::Helpers::Threads::BetaMessageThread)
+        end
+        def thread(
+          max_tokens:,
+          model:,
+          # The messages that seed the history. The first turn sends them.
+          messages: nil,
+          cache_control: nil,
+          compaction: nil,
+          container: nil,
+          context_management: nil,
+          diagnostics: nil,
+          fallback_credit_token: nil,
+          fallbacks: nil,
+          inference_geo: nil,
+          mcp_servers: nil,
+          metadata: nil,
+          output_config: nil,
+          output_format: nil,
+          service_tier: nil,
+          speed: nil,
+          stop_sequences: nil,
+          system_: nil,
+          thinking: nil,
+          tool_choice: nil,
+          tools: nil,
+          betas: nil,
+          user_profile_id: nil,
+          workspace_id: nil,
+          # What a turn does when the API cannot serve the request on a thread: `:error`
+          # (the default) raises the error, and `:continue_unthreaded` sends the turn again,
+          # and every later turn, with the whole history and no thread.
+          unsupported_request_behavior: :error,
+          request_options: {}
+        )
+        end
+
         # See {Anthropic::Resources::Beta::Messages#stream_raw} for streaming counterpart.
         #
         # Send a structured list of input messages with text and/or image content, and the
@@ -316,7 +446,14 @@ module Anthropic
         sig do
           params(
             max_tokens: Integer,
-            messages: T::Array[Anthropic::Beta::BetaMessageParam::OrHash],
+            messages:
+              T::Array[
+                T.any(
+                  Anthropic::Beta::BetaUserMessageParam::OrHash,
+                  Anthropic::Beta::BetaAssistantMessageParam::OrHash,
+                  Anthropic::Beta::BetaSystemMessageParam::OrHash
+                )
+              ],
             model: T.any(Anthropic::Model::OrSymbol, String),
             cache_control:
               T.nilable(Anthropic::Beta::BetaCacheControlEphemeral::OrHash),
@@ -359,7 +496,6 @@ module Anthropic
               T.nilable(Anthropic::Beta::MessageCreateParams::Speed::OrSymbol),
             stop_sequences: T::Array[String],
             system_: Anthropic::Beta::MessageCreateParams::System::Variants,
-            temperature: Float,
             thinking:
               T.any(
                 Anthropic::Beta::BetaThinkingConfigEnabled::OrHash,
@@ -367,6 +503,13 @@ module Anthropic
                 Anthropic::Beta::BetaThinkingConfigBetweenTools::OrHash,
                 Anthropic::Beta::BetaThinkingConfigAdaptive::OrHash
               ),
+            thread:
+              T.nilable(
+                T.any(
+                  Anthropic::Beta::BetaThreadCreateParams::OrHash,
+                  Anthropic::Beta::BetaThreadContinueParams::OrHash
+                )
+              ),
             tool_choice:
               T.any(
                 Anthropic::Beta::BetaToolChoiceAuto::OrHash,
@@ -407,8 +550,6 @@ module Anthropic
                   Anthropic::Beta::BetaMCPToolset::OrHash
                 )
               ],
-            top_k: Integer,
-            top_p: Float,
             betas: T::Array[T.any(Anthropic::AnthropicBeta::OrSymbol, String)],
             user_profile_id: String,
             workspace_id: String,
@@ -496,8 +637,7 @@ module Anthropic
           #
           # Note that if you want to include a
           # [system prompt](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role),
-          # you can use the top-level `system` parameter — there is no `"system"` role for
-          # input messages in the Messages API.
+          # you can use the top-level `system` parameter.
           #
           # There is a limit of 100,000 messages in a single request.
           messages:,
@@ -606,15 +746,6 @@ module Anthropic
           # as specifying a particular goal or role. See our
           # [guide to system prompts](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role).
           system_: nil,
-          # Body param: Amount of randomness injected into the response.
-          #
-          # Defaults to `1.0`. Ranges from `0.0` to `1.0`. Use `temperature` closer to `0.0`
-          # for analytical / multiple choice, and closer to `1.0` for creative and
-          # generative tasks.
-          #
-          # Note that even with `temperature` of `0.0`, the results will not be fully
-          # deterministic.
-          temperature: nil,
           # Body param: Configuration for Claude's thinking.
           #
           # With `{"type": "adaptive"}`, Claude decides when and how much to think. With
@@ -626,6 +757,12 @@ module Anthropic
           # [thinking](https://platform.claude.com/docs/en/build-with-claude/thinking#configuring-thinking)
           # for each model's behavior.
           thinking: nil,
+          # Body param: Server-held conversation state (a Thread). `{"type": "create"}` asks
+          # the server to hold this conversation's state under the response's `id`;
+          # `{"type": "continue", "previous_message_id": ...}` resumes from a previous
+          # response so only the new turn's messages are sent. The response shape is
+          # unchanged.
+          thread: nil,
           # Body param: How the model should use the provided tools. The model can use a
           # specific tool, any available tool, decide by itself, or not use tools at all.
           tool_choice: nil,
@@ -707,26 +844,10 @@ module Anthropic
           # [guide](https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview)
           # for more details.
           tools: nil,
-          # Body param: Only sample from the top K options for each subsequent token.
-          #
-          # Used to remove "long tail" low probability responses.
-          # [Learn more technical details here](https://towardsdatascience.com/how-to-sample-from-language-models-682bceb97277).
-          #
-          # Recommended for advanced use cases only.
-          top_k: nil,
-          # Body param: Use nucleus sampling.
-          #
-          # In nucleus sampling, we compute the cumulative distribution over all the options
-          # for each subsequent token in decreasing probability order and cut it off once it
-          # reaches a particular probability specified by `top_p`.
-          #
-          # Recommended for advanced use cases only.
-          top_p: nil,
           # Header param: Optional header to specify the beta version(s) you want to use.
           betas: nil,
           # Header param: The user profile ID to attribute this request to. Use when acting
-          # on behalf of a party other than your organization. Requires the `user-profiles`
-          # beta header.
+          # on behalf of a party other than your organization.
           user_profile_id: nil,
           # Header param: Optional header to select the Workspace for this request. The
           # value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).
@@ -742,9 +863,6 @@ module Anthropic
         )
         end
 
-        # Hand-written: Anthropic::Resources::Beta::Messages#stream (custom code, not generated).
-        # Its sig and parameters copy #stream_raw's below; update them by hand when those change.
-
         # See {Anthropic::Resources::Beta::Messages#create} for non-streaming counterpart.
         #
         # Send a structured list of input messages with text and/or image content, and the
@@ -758,7 +876,14 @@ module Anthropic
         sig do
           params(
             max_tokens: Integer,
-            messages: T::Array[Anthropic::Beta::BetaMessageParam::OrHash],
+            messages:
+              T::Array[
+                T.any(
+                  Anthropic::Beta::BetaUserMessageParam::OrHash,
+                  Anthropic::Beta::BetaAssistantMessageParam::OrHash,
+                  Anthropic::Beta::BetaSystemMessageParam::OrHash
+                )
+              ],
             model: T.any(Anthropic::Model::OrSymbol, String),
             cache_control:
               T.nilable(Anthropic::Beta::BetaCacheControlEphemeral::OrHash),
@@ -801,7 +926,6 @@ module Anthropic
               T.nilable(Anthropic::Beta::MessageCreateParams::Speed::OrSymbol),
             stop_sequences: T::Array[String],
             system_: Anthropic::Beta::MessageCreateParams::System::Variants,
-            temperature: Float,
             thinking:
               T.any(
                 Anthropic::Beta::BetaThinkingConfigEnabled::OrHash,
@@ -809,6 +933,13 @@ module Anthropic
                 Anthropic::Beta::BetaThinkingConfigBetweenTools::OrHash,
                 Anthropic::Beta::BetaThinkingConfigAdaptive::OrHash
               ),
+            thread:
+              T.nilable(
+                T.any(
+                  Anthropic::Beta::BetaThreadCreateParams::OrHash,
+                  Anthropic::Beta::BetaThreadContinueParams::OrHash
+                )
+              ),
             tool_choice:
               T.any(
                 Anthropic::Beta::BetaToolChoiceAuto::OrHash,
@@ -849,16 +980,19 @@ module Anthropic
                   Anthropic::Beta::BetaMCPToolset::OrHash
                 )
               ],
-            top_k: Integer,
-            top_p: Float,
             betas: T::Array[T.any(Anthropic::AnthropicBeta::OrSymbol, String)],
             user_profile_id: String,
             workspace_id: String,
             stream: T.noreturn,
             request_options: Anthropic::RequestOptions::OrHash
-          ).returns(Anthropic::Helpers::Streaming::MessageStream)
+          ).returns(
+            Anthropic::Internal::Type::BaseStream[
+              T.noreturn,
+              Anthropic::Beta::BetaRawMessageStreamEvent::Variants
+            ]
+          )
         end
-        def stream(
+        def stream_raw(
           # Body param: The maximum number of tokens to generate before stopping.
           #
           # Note that our models may stop _before_ reaching this maximum. This parameter
@@ -886,7 +1020,10 @@ module Anthropic
           #
           # If the final message uses the `assistant` role, the response content will
           # continue immediately from the content in that message. This can be used to
-          # constrain part of the model's response.
+          # constrain part of the model's response. This is called prefill. On models that
+          # don't support prefill, creating a message that ends with a partial `assistant`
+          # response returns a 400 error. See
+          # [Prefill not supported](https://platform.claude.com/docs/en/api/errors#prefill-not-supported).
           #
           # Example with a single `user` message:
           #
@@ -904,7 +1041,8 @@ module Anthropic
           # ]
           # ```
           #
-          # Example with a partially-filled response from Claude:
+          # Example with a partially-filled response from Claude, for models that support
+          # prefill:
           #
           # ```json
           # [
@@ -934,8 +1072,7 @@ module Anthropic
           #
           # Note that if you want to include a
           # [system prompt](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role),
-          # you can use the top-level `system` parameter — there is no `"system"` role for
-          # input messages in the Messages API.
+          # you can use the top-level `system` parameter.
           #
           # There is a limit of 100,000 messages in a single request.
           messages:,
@@ -947,14 +1084,15 @@ module Anthropic
           # Body param: Top-level cache control automatically applies a cache_control marker
           # to the last cacheable block in the request.
           cache_control: nil,
-          # Body param: Compact the whole conversation and return a signed `compaction`
-          # block, alone, that a later request sends back first in `messages`, in place of
-          # the messages it summarizes. There is no trigger and no pause flag: sending the
-          # parameter compacts, and nothing is sampled after the block.
-          #
-          # The summarization prompt is the server's own unless `instructions` are given,
-          # which then replace it for this request; a value that is empty or only whitespace
-          # counts as absent.
+          # Body param: Compaction configuration.
+          #
+          # When set on `POST /v1/messages`, the request is a compaction request: the
+          # conversation in `messages` is summarized and the response holds only the
+          # resulting `compaction` block (`stop_reason` `"compaction"`), which later
+          # requests send first in `messages` in place of the messages it summarizes.
+          # `POST /v1/messages/count_tokens` accepts this parameter and ignores it: the
+          # count it returns is for the conversation in `messages` as sent. Cannot be
+          # combined with `context_management`.
           compaction: nil,
           # Body param: Container identifier for reuse across requests.
           container: nil,
@@ -963,8 +1101,9 @@ module Anthropic
           # This allows you to control how Claude manages context across multiple requests,
           # such as whether to clear function results or not.
           context_management: nil,
-          # Body param: Request-level diagnostics. Currently carries the previous response
-          # id for prompt-cache divergence reporting.
+          # Body param: Request-level diagnostics. Supply `previous_message_id` to have the
+          # response include `diagnostics.cache_miss_reason` explaining any prompt-cache
+          # divergence from that prior request.
           diagnostics: nil,
           # Body param: The `fallback_credit_token` from a prior refusal's `stop_details`.
           #
@@ -988,8 +1127,12 @@ module Anthropic
           fallback_credit_token: nil,
           # Body param: Opt-in server-side retry on one or more substitute models when the
           # requested model declines for policy reasons. Tried in order: if the first entry
-          # also declines, the second is tried, and so on. The string "default" requests the
-          # requested model's server-defined default fallback configuration.
+          # also declines, the second is tried, and so on. Some models don't support
+          # fallbacks; on those models, a list of fallback models returns a 400 error. See
+          # [Server-side fallback](https://platform.claude.com/docs/en/build-with-claude/refusals-and-fallback#server-side-fallback).
+          # The string "default" requests the requested model's server-defined default
+          # fallback configuration. On a model that doesn't support fallbacks, the request
+          # runs on the requested model alone, so a declined request stays declined.
           fallbacks: nil,
           # Body param: Specifies the geographic region for inference processing. If not
           # specified, the workspace's `default_inference_geo` is used.
@@ -1014,9 +1157,13 @@ module Anthropic
           # [service-tiers](https://platform.claude.com/docs/en/api/service-tiers) for
           # details.
           service_tier: nil,
-          # Body param: Inference speed mode. `fast` provides significantly faster output
-          # token generation at premium pricing. Not all models support `fast`; invalid
-          # combinations are rejected at create time.
+          # Body param: The inference speed mode for this request. `"fast"` enables high
+          # output-tokens-per-second inference.
+          #
+          # Some models require the `fast-mode-2026-02-01` beta header for any `speed`
+          # value. See
+          # [fast mode](https://platform.claude.com/docs/en/build-with-claude/fast-mode) for
+          # which models.
           speed: nil,
           # Body param: Custom text sequences that will cause the model to stop generating.
           #
@@ -1034,25 +1181,23 @@ module Anthropic
           # as specifying a particular goal or role. See our
           # [guide to system prompts](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role).
           system_: nil,
-          # Body param: Amount of randomness injected into the response.
-          #
-          # Defaults to `1.0`. Ranges from `0.0` to `1.0`. Use `temperature` closer to `0.0`
-          # for analytical / multiple choice, and closer to `1.0` for creative and
-          # generative tasks.
-          #
-          # Note that even with `temperature` of `0.0`, the results will not be fully
-          # deterministic.
-          temperature: nil,
-          # Body param: Configuration for enabling Claude's extended thinking.
+          # Body param: Configuration for Claude's thinking.
           #
-          # When enabled, responses include `thinking` content blocks showing Claude's
-          # thinking process before the final answer. Requires a minimum budget of 1,024
-          # tokens and counts towards your `max_tokens` limit.
+          # With `{"type": "adaptive"}`, Claude decides when and how much to think. With
+          # `{"type": "enabled"}` (manual extended thinking), you set a `budget_tokens` of
+          # at least 1,024. Thinking tokens count toward your `max_tokens` limit.
           #
-          # See
-          # [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking)
-          # for details.
+          # Which `type` values are accepted, and what happens when you omit `thinking`,
+          # depend on the model. See
+          # [thinking](https://platform.claude.com/docs/en/build-with-claude/thinking#configuring-thinking)
+          # for each model's behavior.
           thinking: nil,
+          # Body param: Server-held conversation state (a Thread). `{"type": "create"}` asks
+          # the server to hold this conversation's state under the response's `id`;
+          # `{"type": "continue", "previous_message_id": ...}` resumes from a previous
+          # response so only the new turn's messages are sent. The response shape is
+          # unchanged.
+          thread: nil,
           # Body param: How the model should use the provided tools. The model can use a
           # specific tool, any available tool, decide by itself, or not use tools at all.
           tool_choice: nil,
@@ -1134,26 +1279,10 @@ module Anthropic
           # [guide](https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview)
           # for more details.
           tools: nil,
-          # Body param: Only sample from the top K options for each subsequent token.
-          #
-          # Used to remove "long tail" low probability responses.
-          # [Learn more technical details here](https://towardsdatascience.com/how-to-sample-from-language-models-682bceb97277).
-          #
-          # Recommended for advanced use cases only.
-          top_k: nil,
-          # Body param: Use nucleus sampling.
-          #
-          # In nucleus sampling, we compute the cumulative distribution over all the options
-          # for each subsequent token in decreasing probability order and cut it off once it
-          # reaches a particular probability specified by `top_p`.
-          #
-          # Recommended for advanced use cases only.
-          top_p: nil,
           # Header param: Optional header to specify the beta version(s) you want to use.
           betas: nil,
           # Header param: The user profile ID to attribute this request to. Use when acting
-          # on behalf of a party other than your organization. Requires the `user-profiles`
-          # beta header.
+          # on behalf of a party other than your organization.
           user_profile_id: nil,
           # Header param: Optional header to select the Workspace for this request. The
           # value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).
@@ -1168,65 +1297,44 @@ module Anthropic
           request_options: {}
         )
         end
-        # End of hand-written Anthropic::Resources::Beta::Messages#stream.
 
-        # See {Anthropic::Resources::Beta::Messages#create} for non-streaming counterpart.
-        #
-        # Send a structured list of input messages with text and/or image content, and the
-        # model will generate the next message in the conversation.
+        # Count the number of tokens in a Message.
         #
-        # The Messages API can be used for either single queries or stateless multi-turn
-        # conversations.
+        # The Token Count API can be used to count the number of tokens in a Message,
+        # including tools, images, and documents, without creating it.
         #
-        # Learn more about the Messages API in our
-        # [user guide](https://platform.claude.com/docs/en/get-started)
+        # Learn more about token counting in our
+        # [user guide](https://platform.claude.com/docs/en/build-with-claude/token-counting)
         sig do
           params(
-            max_tokens: Integer,
-            messages: T::Array[Anthropic::Beta::BetaMessageParam::OrHash],
+            messages:
+              T::Array[
+                T.any(
+                  Anthropic::Beta::BetaUserMessageParam::OrHash,
+                  Anthropic::Beta::BetaAssistantMessageParam::OrHash,
+                  Anthropic::Beta::BetaSystemMessageParam::OrHash
+                )
+              ],
             model: T.any(Anthropic::Model::OrSymbol, String),
             cache_control:
               T.nilable(Anthropic::Beta::BetaCacheControlEphemeral::OrHash),
             compaction:
               T.nilable(Anthropic::Beta::BetaCompactionConfig::OrHash),
-            container:
-              T.nilable(
-                T.any(Anthropic::Beta::BetaContainerParams::OrHash, String)
-              ),
             context_management:
               T.nilable(Anthropic::Beta::BetaContextManagementConfig::OrHash),
-            diagnostics:
-              T.nilable(Anthropic::Beta::BetaDiagnosticsParam::OrHash),
-            fallback_credit_token:
-              T.nilable(
-                T.any(
-                  Anthropic::Beta::BetaFallbackCreditTokenParam::OrHash,
-                  String
-                )
-              ),
-            fallbacks:
-              T.nilable(
-                T.any(
-                  Symbol,
-                  T::Array[Anthropic::Beta::BetaFallbackParam::OrHash]
-                )
-              ),
-            inference_geo: T.nilable(String),
             mcp_servers:
               T::Array[
                 Anthropic::Beta::BetaRequestMCPServerURLDefinition::OrHash
               ],
-            metadata: Anthropic::Beta::BetaMetadata::OrHash,
             output_config: Anthropic::Beta::BetaOutputConfig::OrHash,
             output_format:
               T.nilable(Anthropic::Beta::BetaJSONOutputFormat::OrHash),
-            service_tier:
-              Anthropic::Beta::MessageCreateParams::ServiceTier::OrSymbol,
             speed:
-              T.nilable(Anthropic::Beta::MessageCreateParams::Speed::OrSymbol),
-            stop_sequences: T::Array[String],
-            system_: Anthropic::Beta::MessageCreateParams::System::Variants,
-            temperature: Float,
+              T.nilable(
+                Anthropic::Beta::MessageCountTokensParams::Speed::OrSymbol
+              ),
+            system_:
+              Anthropic::Beta::MessageCountTokensParams::System::Variants,
             thinking:
               T.any(
                 Anthropic::Beta::BetaThinkingConfigEnabled::OrHash,
@@ -1274,34 +1382,13 @@ module Anthropic
                   Anthropic::Beta::BetaMCPToolset::OrHash
                 )
               ],
-            top_k: Integer,
-            top_p: Float,
             betas: T::Array[T.any(Anthropic::AnthropicBeta::OrSymbol, String)],
             user_profile_id: String,
             workspace_id: String,
-            stream: T.noreturn,
             request_options: Anthropic::RequestOptions::OrHash
-          ).returns(
-            Anthropic::Internal::Type::BaseStream[
-              T.noreturn,
-              Anthropic::Beta::BetaRawMessageStreamEvent::Variants
-            ]
-          )
+          ).returns(Anthropic::Beta::BetaMessageTokensCount)
         end
-        def stream_raw(
-          # Body param: The maximum number of tokens to generate before stopping.
-          #
-          # Note that our models may stop _before_ reaching this maximum. This parameter
-          # only specifies the absolute maximum number of tokens to generate.
-          #
-          # Set to `0` to populate the
-          # [prompt cache](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#pre-warming-the-cache)
-          # without generating a response.
-          #
-          # Different models have different maximum values for this parameter. See
-          # [models](https://platform.claude.com/docs/en/about-claude/models/overview) for
-          # details.
-          max_tokens:,
+        def count_tokens(
           # Body param: Input messages.
           #
           # Our models are trained to operate on alternating `user` and `assistant`
@@ -1368,8 +1455,7 @@ module Anthropic
           #
           # Note that if you want to include a
           # [system prompt](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role),
-          # you can use the top-level `system` parameter — there is no `"system"` role for
-          # input messages in the Messages API.
+          # you can use the top-level `system` parameter.
           #
           # There is a limit of 100,000 messages in a single request.
           messages:,
@@ -1391,53 +1477,13 @@ module Anthropic
           # count it returns is for the conversation in `messages` as sent. Cannot be
           # combined with `context_management`.
           compaction: nil,
-          # Body param: Container identifier for reuse across requests.
-          container: nil,
           # Body param: Context management configuration.
           #
           # This allows you to control how Claude manages context across multiple requests,
           # such as whether to clear function results or not.
           context_management: nil,
-          # Body param: Request-level diagnostics. Supply `previous_message_id` to have the
-          # response include `diagnostics.cache_miss_reason` explaining any prompt-cache
-          # divergence from that prior request.
-          diagnostics: nil,
-          # Body param: The `fallback_credit_token` from a prior refusal's `stop_details`.
-          #
-          # When a preceding request was refused and returned a `fallback_credit_token`,
-          # pass that code here on the retry to have the retry's cache-creation tokens for
-          # the prefix that was warm on the refused model billed at the cache-read rate.
-          # Must be redeemed by the same organization and workspace, with the same request
-          # body (optionally extended by one appended `assistant` message whose content is
-          # the partial text — with any trailing whitespace stripped from the final text
-          # block — and paired server-tool blocks streamed before the refusal; the
-          # appended-assistant form is not available for requests with `output_format` set
-          # or forced `tool_choice`), on an eligible fallback model, on the same platform,
-          # and within 5 minutes of the refusal; a mismatch is a 400. A token minted
-          # mid-server-tool-loop whose partial content was continuable may only be redeemed
-          # with the appended-assistant form — if an exact-body retry is rejected with a 400
-          # saying the token must be redeemed by continuing the partial response, retry with
-          # the appended-assistant form instead.
-          #
-          # When the appended-assistant form is used on a model that otherwise disallows
-          # assistant-turn prefill, this token also authorizes that one prefill.
-          fallback_credit_token: nil,
-          # Body param: Opt-in server-side retry on one or more substitute models when the
-          # requested model declines for policy reasons. Tried in order: if the first entry
-          # also declines, the second is tried, and so on. Some models don't support
-          # fallbacks; on those models, a list of fallback models returns a 400 error. See
-          # [Server-side fallback](https://platform.claude.com/docs/en/build-with-claude/refusals-and-fallback#server-side-fallback).
-          # The string "default" requests the requested model's server-defined default
-          # fallback configuration. On a model that doesn't support fallbacks, the request
-          # runs on the requested model alone, so a declined request stays declined.
-          fallbacks: nil,
-          # Body param: Specifies the geographic region for inference processing. If not
-          # specified, the workspace's `default_inference_geo` is used.
-          inference_geo: nil,
           # Body param: MCP servers to be utilized in this request
           mcp_servers: nil,
-          # Body param: An object describing metadata about the request.
-          metadata: nil,
           # Body param: Configuration options for the model's output, such as the output
           # format.
           output_config: nil,
@@ -1447,13 +1493,6 @@ module Anthropic
           # A schema to specify Claude's output format in responses. This parameter will be
           # removed in a future release.
           output_format: nil,
-          # Body param: Determines whether to use priority capacity (if available) or
-          # standard capacity for this request.
-          #
-          # Anthropic offers different levels of service for your API requests. See
-          # [service-tiers](https://platform.claude.com/docs/en/api/service-tiers) for
-          # details.
-          service_tier: nil,
           # Body param: The inference speed mode for this request. `"fast"` enables high
           # output-tokens-per-second inference.
           #
@@ -1462,31 +1501,12 @@ module Anthropic
           # [fast mode](https://platform.claude.com/docs/en/build-with-claude/fast-mode) for
           # which models.
           speed: nil,
-          # Body param: Custom text sequences that will cause the model to stop generating.
-          #
-          # Our models will normally stop when they have naturally completed their turn,
-          # which will result in a response `stop_reason` of `"end_turn"`.
-          #
-          # If you want the model to stop generating when it encounters custom strings of
-          # text, you can use the `stop_sequences` parameter. If the model encounters one of
-          # the custom sequences, the response `stop_reason` value will be `"stop_sequence"`
-          # and the response `stop_sequence` value will contain the matched stop sequence.
-          stop_sequences: nil,
           # Body param: System prompt.
           #
           # A system prompt is a way of providing context and instructions to Claude, such
           # as specifying a particular goal or role. See our
           # [guide to system prompts](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role).
           system_: nil,
-          # Body param: Amount of randomness injected into the response.
-          #
-          # Defaults to `1.0`. Ranges from `0.0` to `1.0`. Use `temperature` closer to `0.0`
-          # for analytical / multiple choice, and closer to `1.0` for creative and
-          # generative tasks.
-          #
-          # Note that even with `temperature` of `0.0`, the results will not be fully
-          # deterministic.
-          temperature: nil,
           # Body param: Configuration for Claude's thinking.
           #
           # With `{"type": "adaptive"}`, Claude decides when and how much to think. With
@@ -1579,26 +1599,10 @@ module Anthropic
           # [guide](https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview)
           # for more details.
           tools: nil,
-          # Body param: Only sample from the top K options for each subsequent token.
-          #
-          # Used to remove "long tail" low probability responses.
-          # [Learn more technical details here](https://towardsdatascience.com/how-to-sample-from-language-models-682bceb97277).
-          #
-          # Recommended for advanced use cases only.
-          top_k: nil,
-          # Body param: Use nucleus sampling.
-          #
-          # In nucleus sampling, we compute the cumulative distribution over all the options
-          # for each subsequent token in decreasing probability order and cut it off once it
-          # reaches a particular probability specified by `top_p`.
-          #
-          # Recommended for advanced use cases only.
-          top_p: nil,
           # Header param: Optional header to specify the beta version(s) you want to use.
           betas: nil,
           # Header param: The user profile ID to attribute this request to. Use when acting
-          # on behalf of a party other than your organization. Requires the `user-profiles`
-          # beta header.
+          # on behalf of a party other than your organization.
           user_profile_id: nil,
           # Header param: Optional header to select the Workspace for this request. The
           # value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).
@@ -1607,43 +1611,65 @@ module Anthropic
           # credential that belongs to a specific Workspace may omit it; if sent, it must
           # match that Workspace.
           workspace_id: nil,
-          # There is no need to provide `stream:`. Instead, use `#stream_raw` or `#create`
-          # for streaming and non-streaming use cases, respectively.
-          stream: true,
           request_options: {}
         )
         end
 
-        # Count the number of tokens in a Message.
-        #
-        # The Token Count API can be used to count the number of tokens in a Message,
-        # including tools, images, and documents, without creating it.
-        #
-        # Learn more about token counting in our
-        # [user guide](https://platform.claude.com/docs/en/build-with-claude/token-counting)
+        # Streams a message like `stream_raw` and returns a MessageStream, which
+        # accumulates the streamed events and parses tool inputs and structured output.
         sig do
           params(
-            messages: T::Array[Anthropic::Beta::BetaMessageParam::OrHash],
+            max_tokens: Integer,
+            messages:
+              T::Array[
+                T.any(
+                  Anthropic::Beta::BetaUserMessageParam::OrHash,
+                  Anthropic::Beta::BetaAssistantMessageParam::OrHash,
+                  Anthropic::Beta::BetaSystemMessageParam::OrHash
+                )
+              ],
             model: T.any(Anthropic::Model::OrSymbol, String),
             cache_control:
               T.nilable(Anthropic::Beta::BetaCacheControlEphemeral::OrHash),
             compaction:
               T.nilable(Anthropic::Beta::BetaCompactionConfig::OrHash),
+            container:
+              T.nilable(
+                T.any(Anthropic::Beta::BetaContainerParams::OrHash, String)
+              ),
             context_management:
               T.nilable(Anthropic::Beta::BetaContextManagementConfig::OrHash),
+            diagnostics:
+              T.nilable(Anthropic::Beta::BetaDiagnosticsParam::OrHash),
+            fallback_credit_token:
+              T.nilable(
+                T.any(
+                  Anthropic::Beta::BetaFallbackCreditTokenParam::OrHash,
+                  String
+                )
+              ),
+            fallbacks:
+              T.nilable(
+                T.any(
+                  Symbol,
+                  T::Array[Anthropic::Beta::BetaFallbackParam::OrHash]
+                )
+              ),
+            inference_geo: T.nilable(String),
             mcp_servers:
               T::Array[
                 Anthropic::Beta::BetaRequestMCPServerURLDefinition::OrHash
               ],
+            metadata: Anthropic::Beta::BetaMetadata::OrHash,
             output_config: Anthropic::Beta::BetaOutputConfig::OrHash,
             output_format:
               T.nilable(Anthropic::Beta::BetaJSONOutputFormat::OrHash),
+            service_tier:
+              Anthropic::Beta::MessageCreateParams::ServiceTier::OrSymbol,
             speed:
-              T.nilable(
-                Anthropic::Beta::MessageCountTokensParams::Speed::OrSymbol
-              ),
-            system_:
-              Anthropic::Beta::MessageCountTokensParams::System::Variants,
+              T.nilable(Anthropic::Beta::MessageCreateParams::Speed::OrSymbol),
+            stop_sequences: T::Array[String],
+            system_: Anthropic::Beta::MessageCreateParams::System::Variants,
             thinking:
               T.any(
                 Anthropic::Beta::BetaThinkingConfigEnabled::OrHash,
@@ -1651,6 +1677,13 @@ module Anthropic
                 Anthropic::Beta::BetaThinkingConfigBetweenTools::OrHash,
                 Anthropic::Beta::BetaThinkingConfigAdaptive::OrHash
               ),
+            thread:
+              T.nilable(
+                T.any(
+                  Anthropic::Beta::BetaThreadCreateParams::OrHash,
+                  Anthropic::Beta::BetaThreadContinueParams::OrHash
+                )
+              ),
             tool_choice:
               T.any(
                 Anthropic::Beta::BetaToolChoiceAuto::OrHash,
@@ -1695,9 +1728,22 @@ module Anthropic
             user_profile_id: String,
             workspace_id: String,
             request_options: Anthropic::RequestOptions::OrHash
-          ).returns(Anthropic::Beta::BetaMessageTokensCount)
+          ).returns(Anthropic::Helpers::Streaming::MessageStream)
         end
-        def count_tokens(
+        def stream(
+          # Body param: The maximum number of tokens to generate before stopping.
+          #
+          # Note that our models may stop _before_ reaching this maximum. This parameter
+          # only specifies the absolute maximum number of tokens to generate.
+          #
+          # Set to `0` to populate the
+          # [prompt cache](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#pre-warming-the-cache)
+          # without generating a response.
+          #
+          # Different models have different maximum values for this parameter. See
+          # [models](https://platform.claude.com/docs/en/about-claude/models/overview) for
+          # details.
+          max_tokens:,
           # Body param: Input messages.
           #
           # Our models are trained to operate on alternating `user` and `assistant`
@@ -1764,8 +1810,7 @@ module Anthropic
           #
           # Note that if you want to include a
           # [system prompt](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role),
-          # you can use the top-level `system` parameter — there is no `"system"` role for
-          # input messages in the Messages API.
+          # you can use the top-level `system` parameter.
           #
           # There is a limit of 100,000 messages in a single request.
           messages:,
@@ -1787,13 +1832,53 @@ module Anthropic
           # count it returns is for the conversation in `messages` as sent. Cannot be
           # combined with `context_management`.
           compaction: nil,
+          # Body param: Container identifier for reuse across requests.
+          container: nil,
           # Body param: Context management configuration.
           #
           # This allows you to control how Claude manages context across multiple requests,
           # such as whether to clear function results or not.
           context_management: nil,
+          # Body param: Request-level diagnostics. Supply `previous_message_id` to have the
+          # response include `diagnostics.cache_miss_reason` explaining any prompt-cache
+          # divergence from that prior request.
+          diagnostics: nil,
+          # Body param: The `fallback_credit_token` from a prior refusal's `stop_details`.
+          #
+          # When a preceding request was refused and returned a `fallback_credit_token`,
+          # pass that code here on the retry to have the retry's cache-creation tokens for
+          # the prefix that was warm on the refused model billed at the cache-read rate.
+          # Must be redeemed by the same organization and workspace, with the same request
+          # body (optionally extended by one appended `assistant` message whose content is
+          # the partial text — with any trailing whitespace stripped from the final text
+          # block — and paired server-tool blocks streamed before the refusal; the
+          # appended-assistant form is not available for requests with `output_format` set
+          # or forced `tool_choice`), on an eligible fallback model, on the same platform,
+          # and within 5 minutes of the refusal; a mismatch is a 400. A token minted
+          # mid-server-tool-loop whose partial content was continuable may only be redeemed
+          # with the appended-assistant form — if an exact-body retry is rejected with a 400
+          # saying the token must be redeemed by continuing the partial response, retry with
+          # the appended-assistant form instead.
+          #
+          # When the appended-assistant form is used on a model that otherwise disallows
+          # assistant-turn prefill, this token also authorizes that one prefill.
+          fallback_credit_token: nil,
+          # Body param: Opt-in server-side retry on one or more substitute models when the
+          # requested model declines for policy reasons. Tried in order: if the first entry
+          # also declines, the second is tried, and so on. Some models don't support
+          # fallbacks; on those models, a list of fallback models returns a 400 error. See
+          # [Server-side fallback](https://platform.claude.com/docs/en/build-with-claude/refusals-and-fallback#server-side-fallback).
+          # The string "default" requests the requested model's server-defined default
+          # fallback configuration. On a model that doesn't support fallbacks, the request
+          # runs on the requested model alone, so a declined request stays declined.
+          fallbacks: nil,
+          # Body param: Specifies the geographic region for inference processing. If not
+          # specified, the workspace's `default_inference_geo` is used.
+          inference_geo: nil,
           # Body param: MCP servers to be utilized in this request
           mcp_servers: nil,
+          # Body param: An object describing metadata about the request.
+          metadata: nil,
           # Body param: Configuration options for the model's output, such as the output
           # format.
           output_config: nil,
@@ -1803,6 +1888,13 @@ module Anthropic
           # A schema to specify Claude's output format in responses. This parameter will be
           # removed in a future release.
           output_format: nil,
+          # Body param: Determines whether to use priority capacity (if available) or
+          # standard capacity for this request.
+          #
+          # Anthropic offers different levels of service for your API requests. See
+          # [service-tiers](https://platform.claude.com/docs/en/api/service-tiers) for
+          # details.
+          service_tier: nil,
           # Body param: The inference speed mode for this request. `"fast"` enables high
           # output-tokens-per-second inference.
           #
@@ -1811,6 +1903,16 @@ module Anthropic
           # [fast mode](https://platform.claude.com/docs/en/build-with-claude/fast-mode) for
           # which models.
           speed: nil,
+          # Body param: Custom text sequences that will cause the model to stop generating.
+          #
+          # Our models will normally stop when they have naturally completed their turn,
+          # which will result in a response `stop_reason` of `"end_turn"`.
+          #
+          # If you want the model to stop generating when it encounters custom strings of
+          # text, you can use the `stop_sequences` parameter. If the model encounters one of
+          # the custom sequences, the response `stop_reason` value will be `"stop_sequence"`
+          # and the response `stop_sequence` value will contain the matched stop sequence.
+          stop_sequences: nil,
           # Body param: System prompt.
           #
           # A system prompt is a way of providing context and instructions to Claude, such
@@ -1828,6 +1930,12 @@ module Anthropic
           # [thinking](https://platform.claude.com/docs/en/build-with-claude/thinking#configuring-thinking)
           # for each model's behavior.
           thinking: nil,
+          # Body param: Server-held conversation state (a Thread). `{"type": "create"}` asks
+          # the server to hold this conversation's state under the response's `id`;
+          # `{"type": "continue", "previous_message_id": ...}` resumes from a previous
+          # response so only the new turn's messages are sent. The response shape is
+          # unchanged.
+          thread: nil,
           # Body param: How the model should use the provided tools. The model can use a
           # specific tool, any available tool, decide by itself, or not use tools at all.
           tool_choice: nil,
@@ -1912,8 +2020,7 @@ module Anthropic
           # Header param: Optional header to specify the beta version(s) you want to use.
           betas: nil,
           # Header param: The user profile ID to attribute this request to. Use when acting
-          # on behalf of a party other than your organization. Requires the `user-profiles`
-          # beta header.
+          # on behalf of a party other than your organization.
           user_profile_id: nil,
           # Header param: Optional header to select the Workspace for this request. The
           # value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).