parallel-web 0.4.2__tar.gz → 0.5.0__tar.gz

{parallel_web-0.4.2 → parallel_web-0.5.0}/.gitignore

@@ -1,4 +1,5 @@
 .prism.log
+.stdy.log
 _dev
 
 __pycache__

parallel_web-0.5.0/.release-please-manifest.json

@@ -0,0 +1,3 @@
+{
+  ".": "0.5.0"
+}

{parallel_web-0.4.2 → parallel_web-0.5.0}/CHANGELOG.md

@@ -1,5 +1,54 @@
 # Changelog
 
+## 0.5.0 (2026-04-21)
+
+Full Changelog: [v0.4.2...v0.5.0](https://github.com/parallel-web/parallel-sdk-python/compare/v0.4.2...v0.5.0)
+
+### Features
+
+* **api:** Add Findall Candidates ([57c4ae2](https://github.com/parallel-web/parallel-sdk-python/commit/57c4ae25d77a8627c6be3673312bfd7f373e17da))
+* **api:** Add Search and Extract v1 and associated types ([ea487f3](https://github.com/parallel-web/parallel-sdk-python/commit/ea487f32b73aee955f662d1fa225841421ce1ba3))
+* **api:** manual - add AdvancedSearchSettings and AdvancedExtractSettings models ([5836a6f](https://github.com/parallel-web/parallel-sdk-python/commit/5836a6fbe8db3a3509556442067f087918edb2bb))
+* **api:** manual updates - update openapi spec ([02db6c0](https://github.com/parallel-web/parallel-sdk-python/commit/02db6c07ec5eb254b18732fcb6f7dc43e31de471))
+* **api:** Remove full_content from OpenAPI Spec ([7a4d651](https://github.com/parallel-web/parallel-sdk-python/commit/7a4d651c3e9f35334175f82daaf6392e9f76dee5))
+* **api:** Search/Extract v1 with advanced_settings and max_results ([4ded29c](https://github.com/parallel-web/parallel-sdk-python/commit/4ded29c2382594f1735101753a0b09a2f7c6972e))
+* **api:** Update OpenAPI spec ([58f19f3](https://github.com/parallel-web/parallel-sdk-python/commit/58f19f38174fb71dc906049c0aef6610ac67971e))
+* **api:** Update OpenAPI spec ([fae95f4](https://github.com/parallel-web/parallel-sdk-python/commit/fae95f4f2c7cb60ebc0babc4fe540617e3334b2d))
+* **internal:** implement indices array format for query and form serialization ([3df5972](https://github.com/parallel-web/parallel-sdk-python/commit/3df5972e34c9aa1709eabc4eb5b8cbbc0adccae2))
+
+
+### Bug Fixes
+
+* **client:** preserve hardcoded query params when merging with user params ([08080bc](https://github.com/parallel-web/parallel-sdk-python/commit/08080bc22c415881cc9f9b05bc22f09ab83c7e8d))
+* **deps:** bump minimum typing-extensions version ([964a46d](https://github.com/parallel-web/parallel-sdk-python/commit/964a46ddfc9ead64e4105e42192a780bc91716b0))
+* ensure file data are only sent as 1 parameter ([1c15cc0](https://github.com/parallel-web/parallel-sdk-python/commit/1c15cc00b1ae223db8e51893ce23b9a2193e3ae7))
+* **pydantic:** do not pass `by_alias` unless set ([f0793c1](https://github.com/parallel-web/parallel-sdk-python/commit/f0793c171465dd57d0fbf82a3bb2281d046f500e))
+* sanitize endpoint path params ([5931597](https://github.com/parallel-web/parallel-sdk-python/commit/59315972d27246485be5cb52671aecaa3aa46253))
+
+
+### Performance Improvements
+
+* **client:** optimize file structure copying in multipart requests ([00e8564](https://github.com/parallel-web/parallel-sdk-python/commit/00e856464a2828693483a704de83ee5d6c4fe19e))
+
+
+### Chores
+
+* **ci:** skip lint on metadata-only changes ([403448c](https://github.com/parallel-web/parallel-sdk-python/commit/403448c7760e70fe0f4b3998a20f048910e91cd6))
+* **internal:** tweak CI branches ([014c802](https://github.com/parallel-web/parallel-sdk-python/commit/014c80287318df0db6207df1579be00c4717f24d))
+* **internal:** update gitignore ([1f4f6b0](https://github.com/parallel-web/parallel-sdk-python/commit/1f4f6b0e5d2a46e1a5457879e937ea5aa551073c))
+* **tests:** bump steady to v0.19.4 ([ebee2e7](https://github.com/parallel-web/parallel-sdk-python/commit/ebee2e761e2a8587cc6aa4c2decfd6310092b039))
+* **tests:** bump steady to v0.19.5 ([2774099](https://github.com/parallel-web/parallel-sdk-python/commit/2774099f753bc0826e9c6b6e9fbb40d4e72e3405))
+* **tests:** bump steady to v0.19.6 ([8e3ee3d](https://github.com/parallel-web/parallel-sdk-python/commit/8e3ee3d04dc149b2bcedb0e4acd92474fafd8d05))
+* **tests:** bump steady to v0.19.7 ([4bcf12e](https://github.com/parallel-web/parallel-sdk-python/commit/4bcf12e670b7997e23ade3d991711fe1ef741e35))
+* **tests:** bump steady to v0.20.1 ([d82ce60](https://github.com/parallel-web/parallel-sdk-python/commit/d82ce601c5687553b0e96990e418289fd8a14e00))
+* **tests:** bump steady to v0.20.2 ([746ca39](https://github.com/parallel-web/parallel-sdk-python/commit/746ca39c749f899dc9137a2f9be5de9aa39210c6))
+* **tests:** bump steady to v0.22.1 ([dec81af](https://github.com/parallel-web/parallel-sdk-python/commit/dec81afb89f46850635daa89e64debe15717d053))
+
+
+### Refactors
+
+* **tests:** switch from prism to steady ([032745e](https://github.com/parallel-web/parallel-sdk-python/commit/032745ea1a03b3d2516b789a28a3c8b8034660d8))
+
 ## 0.4.2 (2026-03-09)
 
 Full Changelog: [v0.4.1...v0.4.2](https://github.com/parallel-web/parallel-sdk-python/compare/v0.4.1...v0.4.2)

{parallel_web-0.4.2 → parallel_web-0.5.0}/CONTRIBUTING.md

@@ -85,7 +85,7 @@ $ pip install ./path-to-wheel-file.whl
 
 ## Running tests
 
-Most tests require you to [set up a mock server](https://github.com/stoplightio/prism) against the OpenAPI spec to run the tests.
+Most tests require you to [set up a mock server](https://github.com/dgellow/steady) against the OpenAPI spec to run the tests.
 
 ```sh
 $ ./scripts/mock

{parallel_web-0.4.2 → parallel_web-0.5.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: parallel-web
-Version: 0.4.2
+Version: 0.5.0
 Summary: The official Python library for the Parallel API
 Project-URL: Homepage, https://github.com/parallel-web/parallel-sdk-python
 Project-URL: Repository, https://github.com/parallel-web/parallel-sdk-python
@@ -27,7 +27,7 @@ Requires-Dist: distro<2,>=1.7.0
 Requires-Dist: httpx<1,>=0.23.0
 Requires-Dist: pydantic<3,>=1.9.0
 Requires-Dist: sniffio
-Requires-Dist: typing-extensions<5,>=4.10
+Requires-Dist: typing-extensions<5,>=4.14
 Provides-Extra: aiohttp
 Requires-Dist: aiohttp; extra == 'aiohttp'
 Requires-Dist: httpx-aiohttp>=0.1.9; extra == 'aiohttp'

{parallel_web-0.4.2 → parallel_web-0.5.0}/api.md

@@ -4,6 +4,30 @@
 from parallel.types import ErrorObject, ErrorResponse, SourcePolicy, Warning
 ```
 
+# Parallel
+
+Types:
+
+```python
+from parallel.types import (
+    AdvancedExtractSettings,
+    AdvancedSearchSettings,
+    ExcerptSettings,
+    ExtractError,
+    ExtractResponse,
+    ExtractResult,
+    FetchPolicy,
+    SearchResult,
+    UsageItem,
+    WebSearchResult,
+)
+```
+
+Methods:
+
+- <code title="post /v1/extract">client.<a href="./src/parallel/_client.py">extract</a>(\*\*<a href="src/parallel/types/client_extract_params.py">params</a>) -> <a href="./src/parallel/types/extract_response.py">ExtractResponse</a></code>
+- <code title="post /v1/search">client.<a href="./src/parallel/_client.py">search</a>(\*\*<a href="src/parallel/types/client_search_params.py">params</a>) -> <a href="./src/parallel/types/search_result.py">SearchResult</a></code>
+
 # TaskRun
 
 Types:

{parallel_web-0.4.2 → parallel_web-0.5.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "parallel-web"
-version = "0.4.2"
+version = "0.5.0"
 description = "The official Python library for the Parallel API"
 dynamic = ["readme"]
 license = "MIT"
@@ -11,7 +11,7 @@ authors = [
 dependencies = [
   "httpx>=0.23.0, <1",
   "pydantic>=1.9.0, <3",
-  "typing-extensions>=4.10, <5",
+  "typing-extensions>=4.14, <5",
   "anyio>=3.5.0, <5",
   "distro>=1.7.0, <2",
   "sniffio",

{parallel_web-0.4.2 → parallel_web-0.5.0}/src/parallel/_base_client.py

@@ -540,6 +540,10 @@ class BaseClient(Generic[_HttpxClientT, _DefaultStreamT]):
                 files = cast(HttpxRequestFiles, ForceMultipartDict())
 
         prepared_url = self._prepare_url(options.url)
+        # preserve hard-coded query params from the url
+        if params and prepared_url.query:
+            params = {**dict(prepared_url.params.items()), **params}
+            prepared_url = prepared_url.copy_with(raw_path=prepared_url.raw_path.split(b"?", 1)[0])
         if "_" in prepared_url.host:
             # work around https://github.com/encode/httpx/discussions/2880
             kwargs["extensions"] = {"sni_hostname": prepared_url.host.replace("_", "-")}

{parallel_web-0.4.2 → parallel_web-0.5.0}/src/parallel/_client.py

@@ -3,32 +3,54 @@
 from __future__ import annotations
 
 import os
-from typing import TYPE_CHECKING, Any, Mapping
-from typing_extensions import Self, override
+from typing import TYPE_CHECKING, Any, Mapping, Optional
+from typing_extensions import Self, Literal, override
 
 import httpx
 
 from . import _exceptions
 from ._qs import Querystring
+from .types import client_search_params, client_extract_params
 from ._types import (
+    Body,
     Omit,
+    Query,
+    Headers,
     Timeout,
     NotGiven,
     Transport,
     ProxiesTypes,
     RequestOptions,
+    SequenceNotStr,
+    omit,
     not_given,
 )
-from ._utils import is_given, get_async_library
+from ._utils import (
+    is_given,
+    maybe_transform,
+    get_async_library,
+    async_maybe_transform,
+)
 from ._compat import cached_property
 from ._version import __version__
+from ._response import (
+    to_raw_response_wrapper,
+    to_streamed_response_wrapper,
+    async_to_raw_response_wrapper,
+    async_to_streamed_response_wrapper,
+)
 from ._streaming import Stream as Stream, AsyncStream as AsyncStream
 from ._exceptions import ParallelError, APIStatusError
 from ._base_client import (
     DEFAULT_MAX_RETRIES,
     SyncAPIClient,
     AsyncAPIClient,
+    make_request_options,
 )
+from .types.search_result import SearchResult
+from .types.extract_response import ExtractResponse
+from .types.advanced_search_settings_param import AdvancedSearchSettingsParam
+from .types.advanced_extract_settings_param import AdvancedExtractSettingsParam
 
 if TYPE_CHECKING:
     from .resources import beta, task_run
@@ -108,6 +130,13 @@ class Parallel(SyncAPIClient):
 
         Clients submit a natural-language objective with an optional input schema; the service plans retrieval, fetches relevant URLs, and returns outputs that conform to a provided or inferred JSON schema. Supports deep research style queries and can return rich structured JSON outputs. Processors trade-off between cost, latency, and quality. Each processor supports calibrated confidences.
         - Output metadata: citations, excerpts, reasoning, and confidence per field
+
+        Task Groups enable batch execution of many independent Task runs with group-level monitoring and failure handling.
+         - Submit hundreds or thousands of Tasks as a single group
+        - Observe group progress and receive results as they complete
+        - Real-time updates via Server-Sent Events (SSE)
+        - Add tasks to an existing group while it is running
+        - Group-level retry and error aggregation
         """
         from .resources.task_run import TaskRunResource
 
@@ -198,6 +227,161 @@ class Parallel(SyncAPIClient):
     # client.with_options(timeout=10).foo.create(...)
     with_options = copy
 
+    def extract(
+        self,
+        *,
+        urls: SequenceNotStr[str],
+        advanced_settings: Optional[AdvancedExtractSettingsParam] | Omit = omit,
+        client_model: Optional[str] | Omit = omit,
+        max_chars_total: Optional[int] | Omit = omit,
+        objective: Optional[str] | Omit = omit,
+        search_queries: Optional[SequenceNotStr[str]] | Omit = omit,
+        session_id: Optional[str] | Omit = omit,
+        # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
+        # The extra values given here take precedence over values defined on the client or passed to this method.
+        extra_headers: Headers | None = None,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = not_given,
+    ) -> ExtractResponse:
+        """
+        Extracts relevant content from specific web URLs.
+
+        The legacy Extract API reference is available
+        [here](https://docs.parallel.ai/api-reference/legacy/extract-beta/extract).
+
+        Args:
+          urls: URLs to extract content from. Up to 20 URLs.
+
+          advanced_settings: Advanced extract configuration.
+
+              These settings may impact result quality and latency unless used carefully. See
+              https://docs.parallel.ai/search/advanced-extract-settings for more info.
+
+          client_model: The model generating this request and consuming the results. Enables
+              optimizations and tailors default settings for the model's capabilities.
+
+          max_chars_total: Upper bound on total characters across excerpts from all extracted results.
+
+          objective: As in SearchRequest, a natural-language description of the underlying question
+              or goal driving the request. Used together with search_queries to focus excerpts
+              on the most relevant content.
+
+          search_queries: Optional keyword search queries, as in SearchRequest. Used together with
+              objective to focus excerpts on the most relevant content.
+
+          session_id: Session identifier to track calls across separate search and extract calls, to
+              be used as part of a larger task. Specifying it may give better contextual
+              results for subsequent API calls.
+
+          extra_headers: Send extra headers
+
+          extra_query: Add additional query parameters to the request
+
+          extra_body: Add additional JSON properties to the request
+
+          timeout: Override the client-level default timeout for this request, in seconds
+        """
+        return self.post(
+            "/v1/extract",
+            body=maybe_transform(
+                {
+                    "urls": urls,
+                    "advanced_settings": advanced_settings,
+                    "client_model": client_model,
+                    "max_chars_total": max_chars_total,
+                    "objective": objective,
+                    "search_queries": search_queries,
+                    "session_id": session_id,
+                },
+                client_extract_params.ClientExtractParams,
+            ),
+            options=make_request_options(
+                extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout
+            ),
+            cast_to=ExtractResponse,
+        )
+
+    def search(
+        self,
+        *,
+        search_queries: SequenceNotStr[str],
+        advanced_settings: Optional[AdvancedSearchSettingsParam] | Omit = omit,
+        client_model: Optional[str] | Omit = omit,
+        max_chars_total: Optional[int] | Omit = omit,
+        mode: Optional[Literal["basic", "advanced"]] | Omit = omit,
+        objective: Optional[str] | Omit = omit,
+        session_id: Optional[str] | Omit = omit,
+        # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
+        # The extra values given here take precedence over values defined on the client or passed to this method.
+        extra_headers: Headers | None = None,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = not_given,
+    ) -> SearchResult:
+        """
+        Searches the web.
+
+        The legacy Search API reference is available
+        [here](https://docs.parallel.ai/api-reference/legacy/search-beta/search).
+
+        Args:
+          search_queries: Concise keyword search queries, 3-6 words each. At least one query is required,
+              provide 2-3 for best results. Used together with objective to focus results on
+              the most relevant content.
+
+          advanced_settings: Advanced search configuration.
+
+              These settings may impact result quality and latency unless used carefully. See
+              https://docs.parallel.ai/search/advanced-search-settings for more info.
+
+          client_model: The model generating this request and consuming the results. Enables
+              optimizations and tailors default settings for the model's capabilities.
+
+          max_chars_total: Upper bound on total characters across excerpts from all results.
+
+          mode: Search mode preset: supported values are `basic` and `advanced`. Basic mode
+              offers the lowest latency and works best with 2-3 high-quality search_queries.
+              Advanced mode provides higher quality with more advanced retrieval and
+              compression. Defaults to `advanced` when omitted.
+
+          objective: Natural-language description of the underlying question or goal driving the
+              search. Used together with search_queries to focus results on the most relevant
+              content. Should be self-contained with enough context to understand the intent
+              of the search.
+
+          session_id: Session identifier to track calls across separate search and extract calls, to
+              be used as part of a larger task. Specifying it may give better contextual
+              results for subsequent API calls.
+
+          extra_headers: Send extra headers
+
+          extra_query: Add additional query parameters to the request
+
+          extra_body: Add additional JSON properties to the request
+
+          timeout: Override the client-level default timeout for this request, in seconds
+        """
+        return self.post(
+            "/v1/search",
+            body=maybe_transform(
+                {
+                    "search_queries": search_queries,
+                    "advanced_settings": advanced_settings,
+                    "client_model": client_model,
+                    "max_chars_total": max_chars_total,
+                    "mode": mode,
+                    "objective": objective,
+                    "session_id": session_id,
+                },
+                client_search_params.ClientSearchParams,
+            ),
+            options=make_request_options(
+                extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout
+            ),
+            cast_to=SearchResult,
+        )
+
     @override
     def _make_status_error(
         self,
@@ -293,6 +477,13 @@ class AsyncParallel(AsyncAPIClient):
 
         Clients submit a natural-language objective with an optional input schema; the service plans retrieval, fetches relevant URLs, and returns outputs that conform to a provided or inferred JSON schema. Supports deep research style queries and can return rich structured JSON outputs. Processors trade-off between cost, latency, and quality. Each processor supports calibrated confidences.
         - Output metadata: citations, excerpts, reasoning, and confidence per field
+
+        Task Groups enable batch execution of many independent Task runs with group-level monitoring and failure handling.
+         - Submit hundreds or thousands of Tasks as a single group
+        - Observe group progress and receive results as they complete
+        - Real-time updates via Server-Sent Events (SSE)
+        - Add tasks to an existing group while it is running
+        - Group-level retry and error aggregation
         """
         from .resources.task_run import AsyncTaskRunResource
 
@@ -383,6 +574,161 @@ class AsyncParallel(AsyncAPIClient):
     # client.with_options(timeout=10).foo.create(...)
     with_options = copy
 
+    async def extract(
+        self,
+        *,
+        urls: SequenceNotStr[str],
+        advanced_settings: Optional[AdvancedExtractSettingsParam] | Omit = omit,
+        client_model: Optional[str] | Omit = omit,
+        max_chars_total: Optional[int] | Omit = omit,
+        objective: Optional[str] | Omit = omit,
+        search_queries: Optional[SequenceNotStr[str]] | Omit = omit,
+        session_id: Optional[str] | Omit = omit,
+        # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
+        # The extra values given here take precedence over values defined on the client or passed to this method.
+        extra_headers: Headers | None = None,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = not_given,
+    ) -> ExtractResponse:
+        """
+        Extracts relevant content from specific web URLs.
+
+        The legacy Extract API reference is available
+        [here](https://docs.parallel.ai/api-reference/legacy/extract-beta/extract).
+
+        Args:
+          urls: URLs to extract content from. Up to 20 URLs.
+
+          advanced_settings: Advanced extract configuration.
+
+              These settings may impact result quality and latency unless used carefully. See
+              https://docs.parallel.ai/search/advanced-extract-settings for more info.
+
+          client_model: The model generating this request and consuming the results. Enables
+              optimizations and tailors default settings for the model's capabilities.
+
+          max_chars_total: Upper bound on total characters across excerpts from all extracted results.
+
+          objective: As in SearchRequest, a natural-language description of the underlying question
+              or goal driving the request. Used together with search_queries to focus excerpts
+              on the most relevant content.
+
+          search_queries: Optional keyword search queries, as in SearchRequest. Used together with
+              objective to focus excerpts on the most relevant content.
+
+          session_id: Session identifier to track calls across separate search and extract calls, to
+              be used as part of a larger task. Specifying it may give better contextual
+              results for subsequent API calls.
+
+          extra_headers: Send extra headers
+
+          extra_query: Add additional query parameters to the request
+
+          extra_body: Add additional JSON properties to the request
+
+          timeout: Override the client-level default timeout for this request, in seconds
+        """
+        return await self.post(
+            "/v1/extract",
+            body=await async_maybe_transform(
+                {
+                    "urls": urls,
+                    "advanced_settings": advanced_settings,
+                    "client_model": client_model,
+                    "max_chars_total": max_chars_total,
+                    "objective": objective,
+                    "search_queries": search_queries,
+                    "session_id": session_id,
+                },
+                client_extract_params.ClientExtractParams,
+            ),
+            options=make_request_options(
+                extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout
+            ),
+            cast_to=ExtractResponse,
+        )
+
+    async def search(
+        self,
+        *,
+        search_queries: SequenceNotStr[str],
+        advanced_settings: Optional[AdvancedSearchSettingsParam] | Omit = omit,
+        client_model: Optional[str] | Omit = omit,
+        max_chars_total: Optional[int] | Omit = omit,
+        mode: Optional[Literal["basic", "advanced"]] | Omit = omit,
+        objective: Optional[str] | Omit = omit,
+        session_id: Optional[str] | Omit = omit,
+        # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
+        # The extra values given here take precedence over values defined on the client or passed to this method.
+        extra_headers: Headers | None = None,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = not_given,
+    ) -> SearchResult:
+        """
+        Searches the web.
+
+        The legacy Search API reference is available
+        [here](https://docs.parallel.ai/api-reference/legacy/search-beta/search).
+
+        Args:
+          search_queries: Concise keyword search queries, 3-6 words each. At least one query is required,
+              provide 2-3 for best results. Used together with objective to focus results on
+              the most relevant content.
+
+          advanced_settings: Advanced search configuration.
+
+              These settings may impact result quality and latency unless used carefully. See
+              https://docs.parallel.ai/search/advanced-search-settings for more info.
+
+          client_model: The model generating this request and consuming the results. Enables
+              optimizations and tailors default settings for the model's capabilities.
+
+          max_chars_total: Upper bound on total characters across excerpts from all results.
+
+          mode: Search mode preset: supported values are `basic` and `advanced`. Basic mode
+              offers the lowest latency and works best with 2-3 high-quality search_queries.
+              Advanced mode provides higher quality with more advanced retrieval and
+              compression. Defaults to `advanced` when omitted.
+
+          objective: Natural-language description of the underlying question or goal driving the
+              search. Used together with search_queries to focus results on the most relevant
+              content. Should be self-contained with enough context to understand the intent
+              of the search.
+
+          session_id: Session identifier to track calls across separate search and extract calls, to
+              be used as part of a larger task. Specifying it may give better contextual
+              results for subsequent API calls.
+
+          extra_headers: Send extra headers
+
+          extra_query: Add additional query parameters to the request
+
+          extra_body: Add additional JSON properties to the request
+
+          timeout: Override the client-level default timeout for this request, in seconds
+        """
+        return await self.post(
+            "/v1/search",
+            body=await async_maybe_transform(
+                {
+                    "search_queries": search_queries,
+                    "advanced_settings": advanced_settings,
+                    "client_model": client_model,
+                    "max_chars_total": max_chars_total,
+                    "mode": mode,
+                    "objective": objective,
+                    "session_id": session_id,
+                },
+                client_search_params.ClientSearchParams,
+            ),
+            options=make_request_options(
+                extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout
+            ),
+            cast_to=SearchResult,
+        )
+
     @override
     def _make_status_error(
         self,
@@ -423,12 +769,26 @@ class ParallelWithRawResponse:
     def __init__(self, client: Parallel) -> None:
         self._client = client
 
+        self.extract = to_raw_response_wrapper(
+            client.extract,
+        )
+        self.search = to_raw_response_wrapper(
+            client.search,
+        )
+
     @cached_property
     def task_run(self) -> task_run.TaskRunResourceWithRawResponse:
         """The Task API executes web research and extraction tasks.
 
         Clients submit a natural-language objective with an optional input schema; the service plans retrieval, fetches relevant URLs, and returns outputs that conform to a provided or inferred JSON schema. Supports deep research style queries and can return rich structured JSON outputs. Processors trade-off between cost, latency, and quality. Each processor supports calibrated confidences.
         - Output metadata: citations, excerpts, reasoning, and confidence per field
+
+        Task Groups enable batch execution of many independent Task runs with group-level monitoring and failure handling.
+         - Submit hundreds or thousands of Tasks as a single group
+        - Observe group progress and receive results as they complete
+        - Real-time updates via Server-Sent Events (SSE)
+        - Add tasks to an existing group while it is running
+        - Group-level retry and error aggregation
         """
         from .resources.task_run import TaskRunResourceWithRawResponse
 
@@ -447,12 +807,26 @@ class AsyncParallelWithRawResponse:
     def __init__(self, client: AsyncParallel) -> None:
         self._client = client
 
+        self.extract = async_to_raw_response_wrapper(
+            client.extract,
+        )
+        self.search = async_to_raw_response_wrapper(
+            client.search,
+        )
+
     @cached_property
     def task_run(self) -> task_run.AsyncTaskRunResourceWithRawResponse:
         """The Task API executes web research and extraction tasks.
 
         Clients submit a natural-language objective with an optional input schema; the service plans retrieval, fetches relevant URLs, and returns outputs that conform to a provided or inferred JSON schema. Supports deep research style queries and can return rich structured JSON outputs. Processors trade-off between cost, latency, and quality. Each processor supports calibrated confidences.
         - Output metadata: citations, excerpts, reasoning, and confidence per field
+
+        Task Groups enable batch execution of many independent Task runs with group-level monitoring and failure handling.
+         - Submit hundreds or thousands of Tasks as a single group
+        - Observe group progress and receive results as they complete
+        - Real-time updates via Server-Sent Events (SSE)
+        - Add tasks to an existing group while it is running
+        - Group-level retry and error aggregation
         """
         from .resources.task_run import AsyncTaskRunResourceWithRawResponse
 
@@ -471,12 +845,26 @@ class ParallelWithStreamedResponse:
     def __init__(self, client: Parallel) -> None:
         self._client = client
 
+        self.extract = to_streamed_response_wrapper(
+            client.extract,
+        )
+        self.search = to_streamed_response_wrapper(
+            client.search,
+        )
+
     @cached_property
     def task_run(self) -> task_run.TaskRunResourceWithStreamingResponse:
         """The Task API executes web research and extraction tasks.
 
         Clients submit a natural-language objective with an optional input schema; the service plans retrieval, fetches relevant URLs, and returns outputs that conform to a provided or inferred JSON schema. Supports deep research style queries and can return rich structured JSON outputs. Processors trade-off between cost, latency, and quality. Each processor supports calibrated confidences.
         - Output metadata: citations, excerpts, reasoning, and confidence per field
+
+        Task Groups enable batch execution of many independent Task runs with group-level monitoring and failure handling.
+         - Submit hundreds or thousands of Tasks as a single group
+        - Observe group progress and receive results as they complete
+        - Real-time updates via Server-Sent Events (SSE)
+        - Add tasks to an existing group while it is running
+        - Group-level retry and error aggregation
         """
         from .resources.task_run import TaskRunResourceWithStreamingResponse
 
@@ -495,12 +883,26 @@ class AsyncParallelWithStreamedResponse:
     def __init__(self, client: AsyncParallel) -> None:
         self._client = client
 
+        self.extract = async_to_streamed_response_wrapper(
+            client.extract,
+        )
+        self.search = async_to_streamed_response_wrapper(
+            client.search,
+        )
+
     @cached_property
     def task_run(self) -> task_run.AsyncTaskRunResourceWithStreamingResponse:
         """The Task API executes web research and extraction tasks.
 
         Clients submit a natural-language objective with an optional input schema; the service plans retrieval, fetches relevant URLs, and returns outputs that conform to a provided or inferred JSON schema. Supports deep research style queries and can return rich structured JSON outputs. Processors trade-off between cost, latency, and quality. Each processor supports calibrated confidences.
         - Output metadata: citations, excerpts, reasoning, and confidence per field
+
+        Task Groups enable batch execution of many independent Task runs with group-level monitoring and failure handling.
+         - Submit hundreds or thousands of Tasks as a single group
+        - Observe group progress and receive results as they complete
+        - Real-time updates via Server-Sent Events (SSE)
+        - Add tasks to an existing group while it is running
+        - Group-level retry and error aggregation
         """
         from .resources.task_run import AsyncTaskRunResourceWithStreamingResponse