clairic 0.1.0__tar.gz

clairic-0.1.0/.gitignore

@@ -0,0 +1,22 @@
+# Python-generated files
+__pycache__/
+*.py[oc]
+build/
+dist/
+wheels/
+*.egg-info
+
+# Virtual environments
+.venv
+
+# Environment variables
+.env
+
+# Tooling caches
+.pytest_cache/
+.ruff_cache/
+
+# Local agent notes and skills
+.cursor/brainstorms/
+.cursor/plans/
+.claude/skills/

clairic-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Vianney Mixtur
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

clairic-0.1.0/NOTICE

@@ -0,0 +1,18 @@
+clairic is an unofficial project. It is not affiliated with, endorsed by, or
+supported by Cloudflare, Inc. "Cloudflare", "Workers AI" and "Clairic" are
+names of Cloudflare, Inc. products.
+
+Docstrings in this package adapt text from the Cloudflare documentation,
+which Cloudflare publishes under the Creative Commons Attribution 4.0
+International license (CC BY 4.0, https://creativecommons.org/licenses/by/4.0/).
+The adapted text was edited for length and context. Sources:
+
+- Clairic model reference
+  https://developers.cloudflare.com/workers-ai/models/clairic/
+- Clairic input and output JSON schemas
+  https://developers.cloudflare.com/workers-ai/models/clairic/schema-input.json
+  https://developers.cloudflare.com/workers-ai/models/clairic/schema-output.json
+- Extraction models concepts
+  https://developers.cloudflare.com/workers-ai/concepts/extraction-models/
+- "Clairic" launch post, The Cloudflare Blog
+  https://blog.cloudflare.com/clairic-extraction-model/

clairic-0.1.0/PKG-INFO

@@ -0,0 +1,209 @@
+Metadata-Version: 2.5
+Name: clairic
+Version: 0.1.0
+Summary: Unofficial typed Python SDK for Cloudflare's Clairic extraction model, built on Pydantic.
+Project-URL: Homepage, https://github.com/VianneyMI/clairic
+Project-URL: Repository, https://github.com/VianneyMI/clairic
+Project-URL: Issues, https://github.com/VianneyMI/clairic/issues
+Project-URL: Clairic model reference, https://developers.cloudflare.com/workers-ai/models/clairic/
+Author-email: Vianney Mixtur <vianney@consio.ai>
+License: MIT License
+        
+        Copyright (c) 2026 Vianney Mixtur
+        
+        Permission is hereby granted, free of charge, to any person obtaining a copy
+        of this software and associated documentation files (the "Software"), to deal
+        in the Software without restriction, including without limitation the rights
+        to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+        copies of the Software, and to permit persons to whom the Software is
+        furnished to do so, subject to the following conditions:
+        
+        The above copyright notice and this permission notice shall be included in all
+        copies or substantial portions of the Software.
+        
+        THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+        IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+        FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+        AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+        LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+        OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+        SOFTWARE.
+License-File: LICENSE
+License-File: NOTICE
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Classifier: Typing :: Typed
+Requires-Python: >=3.10
+Requires-Dist: httpx>=0.27
+Requires-Dist: pydantic>=2.7
+Requires-Dist: typing-extensions>=4.12
+Description-Content-Type: text/markdown
+
+# clairic
+
+**Unofficial** typed Python SDK for [Cloudflare Clairic](https://developers.cloudflare.com/workers-ai/models/clairic/), the Workers AI extraction model. This project is not affiliated with or endorsed by Cloudflare.
+
+Clairic reads a text and returns typed values for the fields you name: extracted, computed, estimated or forecast, each with a confidence score, a probability of absence, or a percentile range. `clairic` lets you declare those fields as a Pydantic model and get the answers back as typed Pydantic objects.
+
+```python
+from typing import Annotated
+
+from pydantic import BaseModel, Field
+
+from clairic import Clairic, DateAnswer, Estimate, EstimateAnswer, Number, NumberAnswer, TextAnswer
+
+
+class InvoiceExtraction(BaseModel):
+    invoice_number: TextAnswer
+    due_date: DateAnswer = Field(description="Issue date plus the net payment days")
+    total: Annotated[NumberAnswer[float], Number(decimals=2, unit="USD")] = Field(
+        description="Amount due including shipping"
+    )
+    po_number: TextAnswer = Field(description="Purchase order number")
+    days_to_pay: Annotated[EstimateAnswer[int], Estimate(min=0, max=120, step=1)] = Field(
+        description="Days after issue until this invoice is paid"
+    )
+
+
+with Clairic() as clairic:
+    result = clairic.extract(
+        "Invoice INV-2291 from Northwind Supply, issued 2026-09-02, net 30. "
+        "4 pallets at $250.00 each, shipping $80.00.",
+        InvoiceExtraction,
+    )
+
+result.data.total.value  # 1080.0
+result.data.total.confidence  # 0.99995
+result.data.po_number.value  # None: the invoice has no PO number
+result.data.days_to_pay.p90  # 37
+```
+
+## Install
+
+```sh
+pip install clairic
+```
+
+Python 3.10 or later. Dependencies: `httpx` and `pydantic` 2.
+
+## Credentials
+
+The client reads `CLOUDFLARE_ACCOUNT_ID` and `CLOUDFLARE_API_TOKEN`, or takes them as arguments:
+
+```python
+clairic = Clairic(account_id="...", api_token="...")
+```
+
+Create the token in the Cloudflare dashboard with access to Workers AI.
+
+## Declaring fields
+
+Each field of the extraction model is an answer type. The answer type sets the Clairic field type; an `Annotated` spec sets its options; `Field(description=...)` gives Clairic the instructions; a Pydantic `alias` sets the field name sent to Clairic.
+
+| Answer type | Clairic type | Spec (options) |
+|---|---|---|
+| `TextAnswer` | `text`: a span of the input | `Text(unit, null_threshold)` |
+| `NumberAnswer[float]`, `NumberAnswer[int]` | `number` (`int` requests whole numbers) | `Number(unit, null_threshold, decimals)` |
+| `DateAnswer`, `TimeAnswer`, `DatetimeAnswer` | `date`, `time`, `datetime` | `Date(null_threshold)`, `Time(...)`, `Datetime(...)` |
+| `DurationAnswer[float]`, `DurationAnswer[int]` | `duration` | `Duration(unit, null_threshold, decimals)`, unit defaults to seconds |
+| `TextListAnswer` | `list` of text spans | `List(unit, decimals)` |
+| `ListAnswer[int]`, `ListAnswer[float]`, `ListAnswer[date]` | `list` of numbers or dates | `List(unit, decimals)` |
+| `EstimateAnswer[int]`, `EstimateAnswer[float]` | `estimate` with number bounds | `Estimate(min, max, step, scale, unit)`, required |
+| `DateEstimateAnswer` | `estimate` with date bounds | `Estimate(min=date, max=date, step="month")`, required |
+| `ForecastAnswer` | `forecast` of a series | `Forecast(min, max, horizon, interval, step, scale, unit)`, required |
+
+Invalid declarations fail before any request: a non-answer annotation, a missing or mismatched spec, or an option Clairic rejects (such as `decimals=7` or `scale="log"` with `min=0`) raises `ModelDefinitionError` or `ValueError` naming the problem.
+
+Set `reference_date` whenever a field depends on today's date or on dates written without a year:
+
+```python
+clairic.extract(text, InvoiceExtraction, reference_date=date(2026, 9, 2))
+```
+
+## Reading answers
+
+Every answer holds a `value` and the metadata Clairic attaches to it. Your editor completes the attributes, and a type checker knows their types.
+
+- **Single values** (`text`, `number`, `date`, `time`, `datetime`, `duration`): `value`, `guess`, `confidence`, `p_null`. `value` is `None` when `p_null` exceeds the field's `null_threshold` (0.5 by default); `guess` still shows the best candidate.
+- **Text**: `span`, end-exclusive offsets into the input: `text[span.start:span.end]`. Text lists carry `spans`.
+- **Number and date lists**: `confidences`, one per item.
+- **Estimates**: `value` (the mean rounded to the `step` grid), `mean`, `median`, `p10`, `p90`, and for number bounds `p_below` and `p_above`, the probability mass outside `[min, max]`.
+- **Forecasts**: per-step lists `value`, `median`, `p10`, `p90`, and `steps` with the full estimate of each step.
+
+The p10 to p90 range is an 80% interval. Cloudflare measures its calibration but does not guarantee it, and intervals can be too narrow on unfamiliar data. Use `confidence`, `p_null` and the interval to decide which answers to act on and which to review.
+
+## Plain values
+
+`values_as` copies each answer value into the same-named field of a plain Pydantic model, ready for the rest of your program:
+
+```python
+class Invoice(BaseModel):
+    invoice_number: str
+    total: float
+    po_number: str | None
+    days_to_pay: int
+
+
+invoice = result.values_as(Invoice)
+```
+
+Pydantic validates the values, so a `None` landing in a field that does not accept `None` raises `ValidationError`.
+
+## Storing results
+
+A `Result` serializes to JSON and back with its answers, token usage and request id:
+
+```python
+stored = result.model_dump_json()
+result = Result[InvoiceExtraction].model_validate_json(stored)
+```
+
+## Async
+
+```python
+from clairic import AsyncClairic
+
+async with AsyncClairic() as clairic:
+    result = await clairic.extract(text, InvoiceExtraction)
+```
+
+## Errors
+
+| Exception | When |
+|---|---|
+| `ModelDefinitionError` | The extraction model cannot be sent to Clairic. Raised before any request. |
+| `ClairicConfigurationError` | The account id or token is missing. |
+| `AuthenticationError` | HTTP 401: invalid token or unknown account. |
+| `InvalidRequestError` | HTTP 400 or 422: Cloudflare rejected the request content. |
+| `ContextWindowExceededError` | HTTP 413: the input exceeds the 65,536-token context window. |
+| `APIError` | Any other error response, or answers that do not fit the model. Base class of the API errors. |
+| `APIConnectionError`, `APITimeoutError` | Cloudflare could not be reached, or did not answer within the timeout (30 seconds by default). |
+
+API errors carry `status_code`, the Cloudflare error `code`, `message` and `request_id`. When Clairic reports a truncated input, the client emits `ClairicTruncationWarning` and sets `result.truncated`.
+
+## Cost
+
+Clairic bills input tokens only ($0.10 per million at the time of writing). Fields that do not fit next to the input in one 8,192-token pass are split across parallel passes, and each pass reads the whole input, so `result.usage.input_tokens` counts the input once per pass.
+
+## Development
+
+```sh
+uv sync
+uv run ruff format --check && uv run ruff check && uv run pyright && uv run pytest
+uv run pytest -m live                         # real API, needs credentials
+uv run python scripts/check_schema_drift.py   # compare Cloudflare's schemas with the vendored ones
+```
+
+`examples/invoice.py` runs the invoice extraction end to end.
+
+## License and attribution
+
+MIT. Docstrings adapt text from the Cloudflare documentation, published under CC BY 4.0; see `NOTICE`.

clairic-0.1.0/README.md

@@ -0,0 +1,160 @@
+# clairic
+
+**Unofficial** typed Python SDK for [Cloudflare Clairic](https://developers.cloudflare.com/workers-ai/models/clairic/), the Workers AI extraction model. This project is not affiliated with or endorsed by Cloudflare.
+
+Clairic reads a text and returns typed values for the fields you name: extracted, computed, estimated or forecast, each with a confidence score, a probability of absence, or a percentile range. `clairic` lets you declare those fields as a Pydantic model and get the answers back as typed Pydantic objects.
+
+```python
+from typing import Annotated
+
+from pydantic import BaseModel, Field
+
+from clairic import Clairic, DateAnswer, Estimate, EstimateAnswer, Number, NumberAnswer, TextAnswer
+
+
+class InvoiceExtraction(BaseModel):
+    invoice_number: TextAnswer
+    due_date: DateAnswer = Field(description="Issue date plus the net payment days")
+    total: Annotated[NumberAnswer[float], Number(decimals=2, unit="USD")] = Field(
+        description="Amount due including shipping"
+    )
+    po_number: TextAnswer = Field(description="Purchase order number")
+    days_to_pay: Annotated[EstimateAnswer[int], Estimate(min=0, max=120, step=1)] = Field(
+        description="Days after issue until this invoice is paid"
+    )
+
+
+with Clairic() as clairic:
+    result = clairic.extract(
+        "Invoice INV-2291 from Northwind Supply, issued 2026-09-02, net 30. "
+        "4 pallets at $250.00 each, shipping $80.00.",
+        InvoiceExtraction,
+    )
+
+result.data.total.value  # 1080.0
+result.data.total.confidence  # 0.99995
+result.data.po_number.value  # None: the invoice has no PO number
+result.data.days_to_pay.p90  # 37
+```
+
+## Install
+
+```sh
+pip install clairic
+```
+
+Python 3.10 or later. Dependencies: `httpx` and `pydantic` 2.
+
+## Credentials
+
+The client reads `CLOUDFLARE_ACCOUNT_ID` and `CLOUDFLARE_API_TOKEN`, or takes them as arguments:
+
+```python
+clairic = Clairic(account_id="...", api_token="...")
+```
+
+Create the token in the Cloudflare dashboard with access to Workers AI.
+
+## Declaring fields
+
+Each field of the extraction model is an answer type. The answer type sets the Clairic field type; an `Annotated` spec sets its options; `Field(description=...)` gives Clairic the instructions; a Pydantic `alias` sets the field name sent to Clairic.
+
+| Answer type | Clairic type | Spec (options) |
+|---|---|---|
+| `TextAnswer` | `text`: a span of the input | `Text(unit, null_threshold)` |
+| `NumberAnswer[float]`, `NumberAnswer[int]` | `number` (`int` requests whole numbers) | `Number(unit, null_threshold, decimals)` |
+| `DateAnswer`, `TimeAnswer`, `DatetimeAnswer` | `date`, `time`, `datetime` | `Date(null_threshold)`, `Time(...)`, `Datetime(...)` |
+| `DurationAnswer[float]`, `DurationAnswer[int]` | `duration` | `Duration(unit, null_threshold, decimals)`, unit defaults to seconds |
+| `TextListAnswer` | `list` of text spans | `List(unit, decimals)` |
+| `ListAnswer[int]`, `ListAnswer[float]`, `ListAnswer[date]` | `list` of numbers or dates | `List(unit, decimals)` |
+| `EstimateAnswer[int]`, `EstimateAnswer[float]` | `estimate` with number bounds | `Estimate(min, max, step, scale, unit)`, required |
+| `DateEstimateAnswer` | `estimate` with date bounds | `Estimate(min=date, max=date, step="month")`, required |
+| `ForecastAnswer` | `forecast` of a series | `Forecast(min, max, horizon, interval, step, scale, unit)`, required |
+
+Invalid declarations fail before any request: a non-answer annotation, a missing or mismatched spec, or an option Clairic rejects (such as `decimals=7` or `scale="log"` with `min=0`) raises `ModelDefinitionError` or `ValueError` naming the problem.
+
+Set `reference_date` whenever a field depends on today's date or on dates written without a year:
+
+```python
+clairic.extract(text, InvoiceExtraction, reference_date=date(2026, 9, 2))
+```
+
+## Reading answers
+
+Every answer holds a `value` and the metadata Clairic attaches to it. Your editor completes the attributes, and a type checker knows their types.
+
+- **Single values** (`text`, `number`, `date`, `time`, `datetime`, `duration`): `value`, `guess`, `confidence`, `p_null`. `value` is `None` when `p_null` exceeds the field's `null_threshold` (0.5 by default); `guess` still shows the best candidate.
+- **Text**: `span`, end-exclusive offsets into the input: `text[span.start:span.end]`. Text lists carry `spans`.
+- **Number and date lists**: `confidences`, one per item.
+- **Estimates**: `value` (the mean rounded to the `step` grid), `mean`, `median`, `p10`, `p90`, and for number bounds `p_below` and `p_above`, the probability mass outside `[min, max]`.
+- **Forecasts**: per-step lists `value`, `median`, `p10`, `p90`, and `steps` with the full estimate of each step.
+
+The p10 to p90 range is an 80% interval. Cloudflare measures its calibration but does not guarantee it, and intervals can be too narrow on unfamiliar data. Use `confidence`, `p_null` and the interval to decide which answers to act on and which to review.
+
+## Plain values
+
+`values_as` copies each answer value into the same-named field of a plain Pydantic model, ready for the rest of your program:
+
+```python
+class Invoice(BaseModel):
+    invoice_number: str
+    total: float
+    po_number: str | None
+    days_to_pay: int
+
+
+invoice = result.values_as(Invoice)
+```
+
+Pydantic validates the values, so a `None` landing in a field that does not accept `None` raises `ValidationError`.
+
+## Storing results
+
+A `Result` serializes to JSON and back with its answers, token usage and request id:
+
+```python
+stored = result.model_dump_json()
+result = Result[InvoiceExtraction].model_validate_json(stored)
+```
+
+## Async
+
+```python
+from clairic import AsyncClairic
+
+async with AsyncClairic() as clairic:
+    result = await clairic.extract(text, InvoiceExtraction)
+```
+
+## Errors
+
+| Exception | When |
+|---|---|
+| `ModelDefinitionError` | The extraction model cannot be sent to Clairic. Raised before any request. |
+| `ClairicConfigurationError` | The account id or token is missing. |
+| `AuthenticationError` | HTTP 401: invalid token or unknown account. |
+| `InvalidRequestError` | HTTP 400 or 422: Cloudflare rejected the request content. |
+| `ContextWindowExceededError` | HTTP 413: the input exceeds the 65,536-token context window. |
+| `APIError` | Any other error response, or answers that do not fit the model. Base class of the API errors. |
+| `APIConnectionError`, `APITimeoutError` | Cloudflare could not be reached, or did not answer within the timeout (30 seconds by default). |
+
+API errors carry `status_code`, the Cloudflare error `code`, `message` and `request_id`. When Clairic reports a truncated input, the client emits `ClairicTruncationWarning` and sets `result.truncated`.
+
+## Cost
+
+Clairic bills input tokens only ($0.10 per million at the time of writing). Fields that do not fit next to the input in one 8,192-token pass are split across parallel passes, and each pass reads the whole input, so `result.usage.input_tokens` counts the input once per pass.
+
+## Development
+
+```sh
+uv sync
+uv run ruff format --check && uv run ruff check && uv run pyright && uv run pytest
+uv run pytest -m live                         # real API, needs credentials
+uv run python scripts/check_schema_drift.py   # compare Cloudflare's schemas with the vendored ones
+```
+
+`examples/invoice.py` runs the invoice extraction end to end.
+
+## License and attribution
+
+MIT. Docstrings adapt text from the Cloudflare documentation, published under CC BY 4.0; see `NOTICE`.

clairic-0.1.0/pyproject.toml

@@ -0,0 +1,83 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "clairic"
+dynamic = ["version"]
+description = "Unofficial typed Python SDK for Cloudflare's Clairic extraction model, built on Pydantic."
+readme = "README.md"
+authors = [{ name = "Vianney Mixtur", email = "vianney@consio.ai" }]
+license = { file = "LICENSE" }
+requires-python = ">=3.10"
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Developers",
+    "License :: OSI Approved :: MIT License",
+    "Operating System :: OS Independent",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
+    "Programming Language :: Python :: 3.14",
+    "Typing :: Typed",
+]
+dependencies = [
+    "httpx>=0.27",
+    "pydantic>=2.7",
+    "typing-extensions>=4.12",
+]
+
+[project.urls]
+Homepage = "https://github.com/VianneyMI/clairic"
+Repository = "https://github.com/VianneyMI/clairic"
+Issues = "https://github.com/VianneyMI/clairic/issues"
+"Clairic model reference" = "https://developers.cloudflare.com/workers-ai/models/clairic/"
+
+[dependency-groups]
+dev = [
+    "anyio>=4.4",
+    "pyright>=1.1.400",
+    "pytest>=8.3",
+    "python-dotenv>=1.0",
+    "ruff>=0.8",
+]
+
+[tool.hatch.version]
+path = "src/clairic/_version.py"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/clairic"]
+
+[tool.hatch.build.targets.sdist]
+include = ["/src", "/LICENSE", "/NOTICE", "/README.md", "/pyproject.toml"]
+
+[tool.ruff]
+line-length = 100
+target-version = "py310"
+
+[tool.ruff.lint]
+select = ["E", "F", "W", "I", "B", "UP", "SIM", "RUF", "D"]
+ignore = [
+    "D100", "D104", "D105", "D107",  # module, package, magic and __init__ docstrings are optional
+    "D203", "D213",                  # use D211 and D212 (Google style)
+]
+
+[tool.ruff.lint.pydocstyle]
+convention = "google"
+
+[tool.ruff.lint.per-file-ignores]
+"tests/**" = ["D"]
+"scripts/**" = ["D"]
+"examples/**" = ["D"]
+
+[tool.pyright]
+include = ["src", "tests", "scripts", "examples"]
+pythonVersion = "3.10"
+typeCheckingMode = "strict"
+
+[tool.pytest.ini_options]
+testpaths = ["tests", "src"]
+markers = ["live: calls the real Cloudflare API (needs CLOUDFLARE_ACCOUNT_ID and CLOUDFLARE_API_TOKEN)"]
+addopts = "-m 'not live' --doctest-modules"

clairic-0.1.0/src/clairic/__init__.py

@@ -0,0 +1,100 @@
+"""Unofficial typed Python SDK for Cloudflare Clairic.
+
+Clairic is a Workers AI extraction model: it reads a text and returns typed values for named
+fields, with confidence scores, probabilities of absence and percentile ranges. This package is
+not affiliated with or endorsed by Cloudflare.
+
+See https://developers.cloudflare.com/workers-ai/models/clairic/ for the model reference.
+"""
+
+from clairic._client import AsyncClairic, Clairic
+from clairic._version import __version__
+from clairic.answers import (
+    Answer,
+    DateAnswer,
+    DateEstimateAnswer,
+    DatetimeAnswer,
+    DurationAnswer,
+    DurationUnit,
+    EstimateAnswer,
+    ForecastAnswer,
+    ForecastStep,
+    ListAnswer,
+    NumberAnswer,
+    Span,
+    TextAnswer,
+    TextListAnswer,
+    TimeAnswer,
+)
+from clairic.errors import (
+    APIConnectionError,
+    APIError,
+    APITimeoutError,
+    AuthenticationError,
+    ClairicConfigurationError,
+    ClairicError,
+    ClairicTruncationWarning,
+    ContextWindowExceededError,
+    InvalidRequestError,
+    ModelDefinitionError,
+)
+from clairic.result import Result, Usage
+from clairic.specs import (
+    Date,
+    Datetime,
+    Duration,
+    Estimate,
+    Forecast,
+    Interval,
+    List,
+    Number,
+    Scale,
+    Spec,
+    Text,
+    Time,
+)
+
+__all__ = [
+    "APIConnectionError",
+    "APIError",
+    "APITimeoutError",
+    "Answer",
+    "AsyncClairic",
+    "AuthenticationError",
+    "Clairic",
+    "ClairicConfigurationError",
+    "ClairicError",
+    "ClairicTruncationWarning",
+    "ContextWindowExceededError",
+    "Date",
+    "DateAnswer",
+    "DateEstimateAnswer",
+    "Datetime",
+    "DatetimeAnswer",
+    "Duration",
+    "DurationAnswer",
+    "DurationUnit",
+    "Estimate",
+    "EstimateAnswer",
+    "Forecast",
+    "ForecastAnswer",
+    "ForecastStep",
+    "Interval",
+    "InvalidRequestError",
+    "List",
+    "ListAnswer",
+    "ModelDefinitionError",
+    "Number",
+    "NumberAnswer",
+    "Result",
+    "Scale",
+    "Span",
+    "Spec",
+    "Text",
+    "TextAnswer",
+    "TextListAnswer",
+    "Time",
+    "TimeAnswer",
+    "Usage",
+    "__version__",
+]