To fix nondeterministic results causing integration testing to sometimes
fail
Also speeds up from 10s to 0.5
---------
Co-authored-by: ccurme <chester.curme@gmail.com>
- use latest models in examples to highlight support
- standardize on using IDs in examples - no more aliases to improve
determinism in future tests
- bump lock
- in integration tests, fix stale casettes and use `MODEL_NAME`
uniformly where possible
- add case for default max tokens for sonnet-4-5 (was missing)
Largely:
- Remove explicit `"Default is x"` since new refs show default inferred
from sig
- Inline code (useful for eventual parsing)
- Fix code block rendering (indentations)
**Description:**
This PR updated links to the latest Anthropic documentation. Changes
include revised links for model overview, tool usage, web search tool,
text editor tool, and more.
**Issue:**
N/A
**Dependencies:**
None
**Twitter handle:**
N/A
- Beta isn't needed for search result tests anymore
- Add TODO for other tests to come back when generally available
- Regenerate remote MCP snapshot after some testing (now the same, but
fresher)
- Bump deps
* Create usage metadata on
[`message_delta`](https://docs.anthropic.com/en/docs/build-with-claude/streaming#event-types)
instead of at the beginning. Consequently, token counts are not included
during streaming but instead at the end. This allows for accurate
reporting of server-side tool usage (important for billing)
* Add some clarifying comments
* Fix some outstanding Pylance warnings
* Remove unnecessary `text` popping in thinking blocks
* Also now correctly reports `input_cache_read`/`input_cache_creation`
as a result
Multiple models were
[retired](https://docs.anthropic.com/en/docs/about-claude/model-deprecations#model-status)
yesterday.
Tests remain broken until we figure out what to do with the legacy
Anthropic LLM integration— currently uses their (legacy) text
completions API, for which there appear to be no remaining supported
models.
Anthropic updated how they report token counts during streaming today.
See changes to `MessageDeltaUsage` in [this
commit](2da00f26c5 (diff-1a396eba0cd9cd8952dcdb58049d3b13f6b7768ead1411888d66e28211f7bfc5)).
It's clean and simple to grab these fields from the final
`message_delta` event. However, some of them are typed as Optional, and
language
[here](e42451ab3f/src/anthropic/lib/streaming/_messages.py (L462))
suggests they may not always be present. So here we take the required
field from the `message_delta` event as we were doing previously, and
ignore the rest.
- Support features from recent update:
https://www.anthropic.com/news/token-saving-updates (mostly adding
support for built-in tools in `bind_tools`
- Add documentation around prompt caching, token-efficient tool use, and
built-in tools.
- Support thinking blocks in core's `convert_to_openai_messages` (pass
through instead of error)
- Ignore thinking blocks in ChatOpenAI (instead of error)
- Support Anthropic-style image blocks in ChatOpenAI
---
Standard integration tests include a `supports_anthropic_inputs`
property which is currently enabled only for tests on `ChatAnthropic`.
This test enforces compatibility with message histories of the form:
```
- system message
- human message
- AI message with tool calls specified only through `tool_use` content blocks
- human message containing `tool_result` and an additional `text` block
```
It additionally checks support for Anthropic-style image inputs if
`supports_image_inputs` is enabled.
Here we change this test, such that if you enable
`supports_anthropic_inputs`:
- You support AI messages with text and `tool_use` content blocks
- You support Anthropic-style image inputs (if `supports_image_inputs`
is enabled)
- You support thinking content blocks.
That is, we add a test case for thinking content blocks, but we also
remove the requirement of handling tool results within HumanMessages
(motivated by existing agent abstractions, which should all return
ToolMessage). We move that requirement to a ChatAnthropic-specific test.
Structured output will currently always raise a BadRequestError when
Claude 3.7 Sonnet's `thinking` is enabled, because we rely on forced
tool use for structured output and this feature is not supported when
`thinking` is enabled.
Here we:
- Emit a warning if `with_structured_output` is called when `thinking`
is enabled.
- Raise `OutputParserException` if no tool calls are generated.
This is arguably preferable to raising an error in all cases.
```python
from langchain_anthropic import ChatAnthropic
from pydantic import BaseModel
class Person(BaseModel):
name: str
age: int
llm = ChatAnthropic(
model="claude-3-7-sonnet-latest",
max_tokens=5_000,
thinking={"type": "enabled", "budget_tokens": 2_000},
)
structured_llm = llm.with_structured_output(Person) # <-- this generates a warning
```
```python
structured_llm.invoke("Alice is 30.") # <-- works
```
```python
structured_llm.invoke("Hello!") # <-- raises OutputParserException
```
Took a "census" of models supported by init_chat_model-- of those that
return model names in response metadata, these were the only two that
had it keyed under `"model"` instead of `"model_name"`.
We currently return string (and therefore no content blocks / citations)
if the response is of the form
```
[
{"text": "a claim", "citations": [...]},
]
```
There are other cases where we do return citations as-is:
```
[
{"text": "a claim", "citations": [...]},
{"text": "some other text"},
{"text": "another claim", "citations": [...]},
]
```
Here we update to return content blocks including citations in the first
case as well.
