Anthropic Messages
Creates a model response given a structured list of input messages using the Anthropic API.
Anthropic Messages
Creates a model response given a structured list of input messages using the Anthropic API.
⚠️ Using our API via a dedicated deployment? Just replace
api.odeus.aiwith your deployment's base URL:<deployment-url>/api/public
Creates a model response for the given chat conversation. This endpoint follows the Anthropic API specification and the requests are sent to AWS Bedrock Anthropic, Google Vertex Anthropic, or Anthropic's own deployments, depending on the selected model.
To use the API you need an API key. Admins can create API keys in the settings.
All parameters from the Anthropic "Create a message" endpoint are supported according to the Anthropic specifications, with the following exception:
model: To see which models are available for your workspace, query the models endpoint:GET /anthropic/{region}/v1/models. The list of available models might differ if you are using your own API keys in Odeus ("Bring-your-own-keys / BYOK", see here for details).
Rate limits
The rate limit for the Messages endpoint is 500 RPM (requests per minute) and 60,000 TPM (tokens per minute). Rate limits are defined at the workspace level - and not at an API key level. Each model has its own rate limit. If you exceed your rate limit, you will receive a 429 Too Many Requests response.
Please note that the rate limits are subject to change, refer to this documentation for the most up-to-date information.
Using Anthropic-compatible libraries
As the request and response format is the same as the Anthropic API, you can use popular libraries like the Anthropic Python library or the Vercel AI SDK to use the Odeus API.
Example using the Anthropic Python library
from anthropic import Anthropic
client = Anthropic(
base_url="https://api.odeus.ai/anthropic/eu/",
api_key="<YOUR_ODEUS_API_KEY>"
)
message = client.messages.create(
model="claude-sonnet-4-6-default",
messages=[
{ "role": "user", "content": "Write a haiku about cats" }
],
max_tokens=1024,
)
print(message.content[0].text)
Example using the Vercel AI SDK in Node.js
import { generateText } from "ai";
import { createAnthropic } from "@ai-sdk/anthropic";
const odeusProvider = createAnthropic({
baseURL: "https://api.odeus.ai/anthropic/eu/v1",
apiKey: "<YOUR_ODEUS_API_KEY>",
});
const result = await generateText({
model: odeusProvider("claude-sonnet-4-6-default"),
prompt: "Write a haiku about cats",
});
console.log(result.text);
Odeus intentionally blocks browser-origin requests to protect your API key and ensure your applications remain secure. For more information, please see our guide on API Key Best Practices.
OpenAPI
openapi: 3.0.0
info:
title: Odeus API
version: 3.0.0
servers:
- url: https://api.odeus.ai
security:
- bearerAuth: []
paths:
/anthropic/{region}/v1/messages:
post:
tags:
- Messages
summary: Create a Message
description: >-
Create a Message.
Send a structured list of input messages with text and/or image content,
and the model will generate the next message in the conversation.
The Messages API can be used for either single queries or stateless
multi-turn conversations.
operationId: messages_post
parameters:
- name: region
in: path
required: true
description: The region of the API to use.
schema:
type: string
enum:
- eu
- us
requestBody:
content:
application/json:
schema:
$ref: '#/components/schemas/CreateMessageParams'
required: true
responses:
'200':
description: Message object.
content:
application/json:
schema:
$ref: '#/components/schemas/Message'
4XX:
description: >-
Error response.
See Anthropic's [errors
documentation](https://platform.claude.com/docs/en/api/errors) for
more details.
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponse'
security:
- bearerAuth: []
components:
schemas:
CreateMessageParams:
additionalProperties: false
example:
max_tokens: 1024
messages:
- content: Write a haiku about cats.
role: user
model: claude-sonnet-4-20250514
properties:
model:
$ref: '#/components/schemas/Model'
messages:
description: >-
Input messages.
Anthropic's models are trained to operate on alternating `user` and
`agent` conversational turns. When creating a new `Message`, you
specify the prior conversational turns with the `messages`
parameter, and the model then generates the next `Message` in the
conversation.
Each input message must be an object with a `role` and `content`.
You can specify a single `user`-role message, or you can include
multiple `user` and `assistant` messages. The first message must
always use the `user` role.
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.
Example with a single `user` message:
```json
[{"role": "user", "content": "Hello, Claude"}]
```
Example with multiple conversational turns:
```json
[
{"role": "user", "content": "Hello there."},
{"role": "assistant", "content": "Hi, I'm Claude. How can I help you?"},
{"role": "user", "content": "Can you explain LLMs in plain English?"},
]
```
Example with a partially-filled response from Claude:
```json
[
{"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},
{"role": "assistant", "content": "The best answer is ("},
]
```
Each input message `content` may be either a single `string` or an
array of content blocks, where each block has a specific `type`.
Using a `string` for `content` is shorthand for an array of one
content block of type `"text"`. The following input messages are
equivalent:
```json
{"role": "user", "content": "Hello, Claude"}
```
```json
{"role": "user", "content": [{"type": "text", "text": "Hello,
Claude"}]}
```
Starting with Claude 3 models, you can also send image content
blocks:
```json
{"role": "user", "content": [
{
"type": "image",
"source": {
"type": "base64",
"media_type": "image/jpeg",
"data": "/9j/4AAQSkZJRg...",
}
},
{"type": "text", "text": "What is in this image?"}
]}
```
We currently support the `base64` source type for images, and the
`image/jpeg`, `image/png`, `image/gif`, and `image/webp` media
types.
See
[examples](https://platform.claude.com/docs/en/api/messages-examples#vision)
for more input examples.
Note that if you want to include a [system
prompt](https://platform.claude.com/docs/en/docs/system-prompts),
you can use the top-level `system` parameter â€" there is no
`"system"` role for input messages in the Messages API.
items:
$ref: '#/components/schemas/InputMessage'
title: Messages
type: array
max_tokens:
description: >-
The maximum number of tokens to generate before stopping.
Note that Anthropic's models may stop _before_ reaching this
maximum. This parameter only specifies the absolute maximum number
of tokens to generate.
Different models have different maximum values for this parameter.
See
[models](https://platform.claude.com/docs/en/docs/models-overview)
for details.
example:
- 1024
minimum: 1
title: Max Tokens
type: integer
stop_sequences:
description: >-
Custom text sequences that will cause the model to stop generating.
Anthropic's 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.
items:
type: string
title: Stop Sequences
type: array
stream:
description: >-
Whether to incrementally stream the response using server-sent
events.
See
[streaming](https://platform.claude.com/docs/en/build-with-claude/streaming)
for details.
title: Stream
type: boolean
system:
anyOf:
- type: string
x-stainless-skip:
- go
- items:
$ref: '#/components/schemas/RequestTextBlock'
type: array
description: >-
System prompt.
A system prompt is a way of providing context and instructions to
Claude, such as specifying a particular goal or role. See
Anthropic's [guide to system
prompts](https://platform.claude.com/docs/en/docs/system-prompts).
example:
- - text: Today's date is 2024-06-01.
type: text
- Today's date is 2023-01-01.
title: System
temperature:
description: >-
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.
example:
- 1
maximum: 1
minimum: 0
title: Temperature
type: number
tool_choice:
description: >-
How the model should use the provided tools. The model can use a
specific tool, any available tool, or decide by itself.
discriminator:
mapping:
any:
$ref: '#/components/schemas/ToolChoiceAny'
auto:
$ref: '#/components/schemas/ToolChoiceAuto'
tool:
$ref: '#/components/schemas/ToolChoiceTool'
propertyName: type
oneOf:
- $ref: '#/components/schemas/ToolChoiceAuto'
- $ref: '#/components/schemas/ToolChoiceAny'
- $ref: '#/components/schemas/ToolChoiceTool'
title: Tool Choice
tools:
description: >-
Definitions of tools that the model may use.
If you include `tools` in your API request, the model may return
`tool_use` content blocks that represent the model's use of those
tools. You can then run those tools using the tool input generated
by the model and then optionally return results back to the model
using `tool_result` content blocks.
Each tool definition includes:
* `name`: Name of the tool.
* `description`: Optional, but strongly-recommended description of
the tool.
* `input_schema`: [JSON schema](https://json-schema.org/) for the
tool `input` shape that the model will produce in `tool_use` output
content blocks.
For example, if you defined `tools` as:
```json
[
{
"name": "get_stock_price",
"description": "Get the current stock price for a given ticker symbol.",
"input_schema": {
"type": "object",
"properties": {
"ticker": {
"type": "string",
"description": "The stock ticker symbol, e.g. AAPL for Apple Inc."
}
},
"required": ["ticker"]
}
}
]
```
And then asked the model "What's the S&P 500 at today?", the model
might produce `tool_use` content blocks in the response like this:
```json
[
{
"type": "tool_use",
"id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",
"name": "get_stock_price",
"input": { "ticker": "^GSPC" }
}
]
```
You might then run your `get_stock_price` tool with `{"ticker":
"^GSPC"}` as an input, and return the following back to the model in
a subsequent `user` message:
```json
[
{
"type": "tool_result",
"tool_use_id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",
"content": "259.75 USD"
}
]
```
Tools can be used for workflows that include running client-side
tools and functions, or more generally whenever you want the model
to produce a particular JSON structure of output.
See Anthropic's
[guide](https://platform.claude.com/docs/en/docs/tool-use) for more
details.
example:
- description: Get the current weather in a given location
input_schema:
properties:
location:
description: The city and state, e.g. San Francisco, CA
type: string
unit:
description: Unit for the output - one of (celsius, fahrenheit)
type: string
required:
- location
type: object
name: get_weather
items:
$ref: '#/components/schemas/Tool'
title: Tools
type: array
top_k:
description: >-
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-write-expert-prompts-for-chatgpt-gpt-4-and-other-language-models-23133dc85550/).
Recommended for advanced use cases only. You usually only need to
use `temperature`.
example:
- 5
minimum: 0
title: Top K
type: integer
top_p:
description: >-
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`.
example:
- 0.7
maximum: 1
minimum: 0
title: Top P
type: number
thinking:
$ref: '#/components/schemas/ThinkingConfig'
required:
- model
- messages
- max_tokens
title: CreateMessageParams
type: object
Message:
example:
- content:
- text: >-
Here is a haiku about cats:\n\nFeline grace and charm,\nPurring
softly by the fire,\nCats reign supreme.
type: text
id: msg_013Zva2CMHLNnXjNJJKqJ2EF
model: claude-3-haiku-20240307
role: assistant
stop_reason: end_turn
stop_sequence: null
type: message
usage:
input_tokens: 14
output_tokens: 35
properties:
id:
description: |-
Unique object identifier.
The format and length of IDs may change over time.
example:
- msg_013Zva2CMHLNnXjNJJKqJ2EF
title: Id
type: string
type:
default: message
description: |-
Object type.
For Messages, this is always `"message"`.
enum:
- message
title: Type
type: string
role:
default: assistant
description: |-
Conversational role of the generated message.
This will always be `"assistant"`.
enum:
- agent
title: Role
type: string
content:
description: >-
Content generated by the model.
This is an array of content blocks, each of which has a `type` that
determines its shape.
Example:
```json
[{"type": "text", "text": "Hi, I'm Claude."}]
```
If the request input `messages` ended with an `agent` turn, then the
response `content` will continue directly from that last turn. You
can use this to constrain the model's output.
For example, if the input `messages` were:
```json
[
{"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},
{"role": "assistant", "content": "The best answer is ("}
]
```
Then the response `content` might be:
```json
[{"type": "text", "text": "B)"}]
```
example:
- - text: Hi! My name is Claude.
type: text
items:
$ref: '#/components/schemas/ContentBlock'
title: Content
type: array
model:
$ref: '#/components/schemas/Model'
stop_reason:
anyOf:
- enum:
- end_turn
- max_tokens
- stop_sequence
- tool_use
type: string
description: >-
The reason that we stopped.
This may be one the following values:
* `"end_turn"`: the model reached a natural stopping point
* `"max_tokens"`: we exceeded the requested `max_tokens` or the
model's maximum
* `"stop_sequence"`: one of your provided custom `stop_sequences`
was generated
* `"tool_use"`: the model invoked one or more tools
In non-streaming mode this value is always non-null. In streaming
mode, it is null in the `message_start` event and non-null
otherwise.
title: Stop Reason
stop_sequence:
anyOf:
- type: string
default: null
description: >-
Which custom stop sequence was generated, if any.
This value will be a non-null string if one of your custom stop
sequences was generated.
title: Stop Sequence
usage:
allOf:
- $ref: '#/components/schemas/Usage'
description: >-
Input and output token counts, representing the underlying cost to
our systems.
Under the hood, the API transforms requests into a format suitable
for the model. The model's output then goes through a parsing stage
before becoming an API response. As a result, the token counts in
`usage` will not match one-to-one with the exact visible content of
an API request or response.
For example, `output_tokens` will be non-zero, even for an empty
string response from Claude.
example:
- input_tokens: 2095
output_tokens: 503
required:
- id
- type
- role
- content
- model
- stop_reason
- stop_sequence
- usage
title: Message
type: object
x-stainless-python-custom-imports:
- from .content_block import ContentBlock as ContentBlock
ErrorResponse:
properties:
type:
default: error
enum:
- error
title: Type
type: string
error:
discriminator:
mapping:
api_error:
$ref: '#/components/schemas/APIError'
authentication_error:
$ref: '#/components/schemas/AuthenticationError'
invalid_request_error:
$ref: '#/components/schemas/InvalidRequestError'
not_found_error:
$ref: '#/components/schemas/NotFoundError'
overloaded_error:
$ref: '#/components/schemas/OverloadedError'
permission_error:
$ref: '#/components/schemas/PermissionError'
rate_limit_error:
$ref: '#/components/schemas/RateLimitError'
propertyName: type
oneOf:
- $ref: '#/components/schemas/InvalidRequestError'
- $ref: '#/components/schemas/AuthenticationError'
- $ref: '#/components/schemas/PermissionError'
- $ref: '#/components/schemas/NotFoundError'
- $ref: '#/components/schemas/RateLimitError'
- $ref: '#/components/schemas/APIError'
- $ref: '#/components/schemas/OverloadedError'
title: Error
required:
- type
- error
title: ErrorResponse
type: object
Model:
title: Model
description: >-
The model that will complete your prompt. See
[models](https://platform.claude.com/docs/en/docs/models-overview) for
additional details and options.
type: string
InputMessage:
additionalProperties: false
properties:
role:
enum:
- user
- agent
title: Role
type: string
content:
anyOf:
- type: string
x-stainless-skip:
- go
- items:
discriminator:
mapping:
image:
$ref: '#/components/schemas/RequestImageBlock'
text:
$ref: '#/components/schemas/RequestTextBlock'
tool_result:
$ref: '#/components/schemas/RequestToolResultBlock'
tool_use:
$ref: '#/components/schemas/RequestToolUseBlock'
propertyName: type
oneOf:
- $ref: '#/components/schemas/RequestTextBlock'
- $ref: '#/components/schemas/RequestImageBlock'
- $ref: '#/components/schemas/RequestToolUseBlock'
- $ref: '#/components/schemas/RequestToolResultBlock'
x-stainless-python-extend-union:
- ContentBlock
x-stainless-python-extend-union-imports:
- from .content_block import ContentBlock
type: array
example:
- type: text
text: What is a quaternion?
title: Content
required:
- role
- content
title: InputMessage
type: object
RequestTextBlock:
additionalProperties: false
properties:
type:
enum:
- text
title: Type
type: string
text:
minLength: 1
title: Text
type: string
required:
- type
- text
title: Text
type: object
ToolChoiceAny:
additionalProperties: false
description: The model will use any available tools.
properties:
type:
enum:
- any
title: Type
type: string
required:
- type
title: ToolChoiceAny
type: object
ToolChoiceAuto:
additionalProperties: false
description: The model will automatically decide whether to use tools.
properties:
type:
enum:
- auto
title: Type
type: string
required:
- type
title: ToolChoiceAuto
type: object
ToolChoiceTool:
additionalProperties: false
description: The model will use the specified tool with `tool_choice.name`.
properties:
type:
enum:
- tool
title: Type
type: string
name:
description: The name of the tool to use.
title: Name
type: string
required:
- type
- name
title: ToolChoiceTool
type: object
Tool:
additionalProperties: false
properties:
description:
description: >-
Description of what this tool does.
Tool descriptions should be as detailed as possible. The more
information that the model has about what the tool is and how to use
it, the better it will perform. You can use natural language
descriptions to reinforce important aspects of the tool input JSON
schema.
example:
- Get the current weather in a given location
title: Description
type: string
name:
maxLength: 64
minLength: 1
pattern: ^[a-zA-Z0-9_-]{1,64}$
title: Name
type: string
input_schema:
allOf:
- $ref: '#/components/schemas/InputSchema'
description: >-
[JSON schema](https://json-schema.org/) for this tool's input.
This defines the shape of the `input` that your tool accepts and
that the model will produce.
example:
- properties:
location:
description: The city and state, e.g. San Francisco, CA
type: string
unit:
description: Unit for the output - one of (celsius, fahrenheit)
type: string
required:
- location
type: object
required:
- name
- input_schema
title: Tool
type: object
ThinkingConfig:
type: object
title: Thinking
description: >-
Configuration for enabling Claude's extended thinking. When enabled,
responses include `thinking` content blocks showing Claude's reasoning
process before the final answer.
See [extended
thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking)
for details.
properties:
type:
type: string
enum:
- enabled
description: Must be "enabled" to activate extended thinking.
budget_tokens:
type: integer
minimum: 1024
description: >-
Maximum number of tokens Claude can use for thinking. Must be ≥1024
and less than max_tokens.
required:
- type
- budget_tokens
ContentBlock:
discriminator:
mapping:
text:
$ref: '#/components/schemas/ResponseTextBlock'
tool_use:
$ref: '#/components/schemas/ResponseToolUseBlock'
thinking:
$ref: '#/components/schemas/ResponseThinkingBlock'
propertyName: type
oneOf:
- $ref: '#/components/schemas/ResponseTextBlock'
- $ref: '#/components/schemas/ResponseToolUseBlock'
- $ref: '#/components/schemas/ResponseThinkingBlock'
Usage:
properties:
input_tokens:
description: The number of input tokens which were used.
example:
- 2095
title: Input Tokens
type: integer
output_tokens:
description: The number of output tokens which were used.
example:
- 503
title: Output Tokens
type: integer
required:
- input_tokens
- output_tokens
title: Usage
type: object
APIError:
properties:
type:
default: api_error
enum:
- api_error
title: Type
type: string
message:
default: Internal server error
title: Message
type: string
required:
- type
- message
title: APIError
type: object
AuthenticationError:
properties:
type:
default: authentication_error
enum:
- authentication_error
title: Type
type: string
message:
default: Authentication error
title: Message
type: string
required:
- type
- message
title: AuthenticationError
type: object
InvalidRequestError:
properties:
type:
default: invalid_request_error
enum:
- invalid_request_error
title: Type
type: string
message:
default: Invalid request
title: Message
type: string
required:
- type
- message
title: InvalidRequestError
type: object
NotFoundError:
properties:
type:
default: not_found_error
enum:
- not_found_error
title: Type
type: string
message:
default: Not found
title: Message
type: string
required:
- type
- message
title: NotFoundError
type: object
OverloadedError:
properties:
type:
default: overloaded_error
enum:
- overloaded_error
title: Type
type: string
message:
default: Overloaded
title: Message
type: string
required:
- type
- message
title: OverloadedError
type: object
PermissionError:
properties:
type:
default: permission_error
enum:
- permission_error
title: Type
type: string
message:
default: Permission denied
title: Message
type: string
required:
- type
- message
title: PermissionError
type: object
RateLimitError:
properties:
type:
default: rate_limit_error
enum:
- rate_limit_error
title: Type
type: string
message:
default: Rate limited
title: Message
type: string
required:
- type
- message
title: RateLimitError
type: object
RequestImageBlock:
additionalProperties: false
properties:
type:
enum:
- image
title: Type
type: string
source:
discriminator:
mapping:
base64:
$ref: '#/components/schemas/Base64ImageSource'
propertyName: type
oneOf:
- $ref: '#/components/schemas/Base64ImageSource'
title: Source
required:
- type
- source
title: Image
type: object
RequestToolResultBlock:
additionalProperties: false
properties:
type:
enum:
- tool_result
title: Type
type: string
tool_use_id:
pattern: ^[a-zA-Z0-9_-]+$
title: Tool Use Id
type: string
is_error:
default: false
title: Is Error
type: boolean
content:
anyOf:
- type: string
x-stainless-skip:
- go
- items:
discriminator:
mapping:
image:
$ref: '#/components/schemas/RequestImageBlock'
text:
$ref: '#/components/schemas/RequestTextBlock'
propertyName: type
oneOf:
- $ref: '#/components/schemas/RequestTextBlock'
- $ref: '#/components/schemas/RequestImageBlock'
type: array
x-stainless-naming:
python:
type_name: Content
title: Content
required:
- type
- tool_use_id
title: Tool Result
type: object
RequestToolUseBlock:
additionalProperties: false
properties:
type:
enum:
- tool_use
title: Type
type: string
id:
pattern: ^[a-zA-Z0-9_-]+$
title: Id
type: string
name:
maxLength: 64
minLength: 1
pattern: ^[a-zA-Z0-9_-]{1,64}$
title: Name
type: string
input:
title: Input
type: object
required:
- type
- id
- name
- input
title: Tool Use
type: object
InputSchema:
additionalProperties: true
properties:
type:
enum:
- object
title: Type
type: string
properties:
anyOf:
- type: object
default: null
title: Properties
required:
- type
title: InputSchema
type: object
ResponseTextBlock:
properties:
type:
default: text
enum:
- text
title: Type
type: string
text:
maxLength: 5000000
minLength: 0
title: Text
type: string
required:
- type
- text
title: Text
type: object
ResponseToolUseBlock:
properties:
type:
default: tool_use
enum:
- tool_use
title: Type
type: string
id:
pattern: ^[a-zA-Z0-9_-]+$
title: Id
type: string
name:
minLength: 1
title: Name
type: string
input:
title: Input
type: object
required:
- type
- id
- name
- input
title: Tool Use
type: object
ResponseThinkingBlock:
type: object
description: >-
A thinking content block containing Claude's reasoning process when
extended thinking is enabled.
properties:
type:
type: string
enum:
- thinking
description: The type of content block.
thinking:
type: string
description: Claude's internal reasoning process.
signature:
type: string
description: A signature verifying the thinking block's authenticity.
required:
- type
- thinking
Base64ImageSource:
additionalProperties: false
properties:
type:
enum:
- base64
title: Type
type: string
media_type:
enum:
- image/jpeg
- image/png
- image/gif
- image/webp
title: Media Type
type: string
data:
format: byte
title: Data
type: string
required:
- type
- media_type
- data
title: Base64ImageSource
type: object
securitySchemes:
bearerAuth:
type: http
scheme: bearer
bearerFormat: API Key
description: API key as Bearer token. Format "Bearer YOUR_API_KEY"