Chat Completions

POST https://api.synthetic.new/openai/v1/chat/completions

Create a model response for a given chat conversation.

Request Body

Parameter	Type	Required	Description
`messages`	array	Yes	List of messages comprising the conversation
`model`	string	Yes	Model name (must be prefixed with `hf:`). See supported Models.
`frequency_penalty`	number	No	Penalty for token frequency (-2.0 to 2.0). Reduces repetition.
`logit_bias`	object	No	Modify token likelihood. Maps token IDs to bias values (-100 to 100).
`logprobs`	boolean/number	No	Return log probabilities of output tokens
`max_completion_tokens`	number	No	Maximum tokens for completion (including reasoning tokens)
`max_tokens`	number	No	Maximum number of tokens to generate
`min_p`	number	No	Minimum probability for nucleus sampling
`n`	number	No	Number of completion choices to generate (default: 1)
`parallel_tool_calls`	boolean	No	Enable parallel function calling during tool use
`presence_penalty`	number	No	Penalty for token presence (-2.0 to 2.0). Encourages new topics.
`reasoning_effort`	string	No	Control reasoning effort for thinking models: `low`, `medium`, or `high`
`response_format`	object	No	Specify output format (JSON schema, JSON object, etc.)
`seed`	number	No	Random seed for deterministic sampling
`stop`	string/array	No	Stop sequence(s)
`stream`	boolean	No	Stream response using server-sent events
`stream_options`	object	No	Options for streaming (when `stream: true`)
`temperature`	number	No	Sampling randomness (0.0-2.0). Higher = more random.
`tool_choice`	string/object	No	Control tool calling: `auto`, `none`, or specific tool
`tools`	array	No	List of tools the model may call
`top_k`	number	No	Limit sampling to top K tokens
`top_logprobs`	number	No	Number of top token probabilities to return
`top_p`	number	No	Nucleus sampling threshold (0.0-1.0)
`user`	string	No	Unique identifier representing your end-user
`function_call`	string/object	No	(Deprecated) Control function calling: `auto`, `none`, or specific function
`functions`	array	No	(Deprecated) List of functions the model may call

Message Object

Parameter	Type	Required	Description
`role`	string	Yes	Role: `system`, `user`, `assistant`, `tool`, or `function`
`content`	string/array	Yes	Message content (text string or array of content parts)
`name`	string	No	Name of the participant (for user/system messages)
`tool_calls`	array	No	Tool calls to be executed (for assistant messages)
`tool_call_id`	string	No	ID of the tool call (for tool messages)
`reasoning_content`	string	No	Reasoning content for thinking models (for assistant messages)
`function_call`	object	No	(Deprecated) Function call to be executed (for assistant messages)

Example Request

Python
TypeScript
curl

import openai

client = openai.OpenAI(
  api_key="SYNTHETIC_API_KEY",
  base_url="https://api.synthetic.new/openai/v1"
)

completion = client.chat.completions.create(
  model="hf:deepseek-ai/DeepSeek-V3-0324",
  messages=[
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "What is the capital of France?"}
  ],
  max_tokens=100,
  temperature=0.7
)

print(completion.choices[0].message.content)

Example Response

json

{
  "id": "296a8fcb-181d-46d4-ad9b-b380befa9a07",
  "object": "chat.completion",
  "created": 1757644754,
  "model": "accounts/fireworks/models/deepseek-v3-0324",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "The capital of France is **Paris**. \n\nParis is known for its rich history, iconic landmarks like the Eiffel Tower and the Louvre Museum, and its cultural significance as a global hub for art, fashion, and cuisine. \n\nLet me know if you'd like more details!"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 16,
    "total_tokens": 74,
    "completion_tokens": 58
  }
}

Streaming

When stream: true is set, the response will be a series of Server-Sent Events:

Python
TypeScript
curl

completion = client.chat.completions.create(
  model="hf:deepseek-ai/DeepSeek-V3-0324",
  messages=[
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Tell me a short story."}
  ],
  max_tokens=100,
  stream=True
)

for chunk in completion:
  if chunk.choices[0].delta.content:
    print(chunk.choices[0].delta.content, end='', flush=True)

Streaming Response

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1757644754,"model":"accounts/fireworks/models/deepseek-v3-0324","choices":[{"index":0,"delta":{"role":"assistant","content":"Once"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1757644754,"model":"accounts/fireworks/models/deepseek-v3-0324","choices":[{"index":0,"delta":{"content":" upon"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1757644754,"model":"accounts/fireworks/models/deepseek-v3-0324","choices":[{"index":0,"delta":{"content":" a"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1757644754,"model":"accounts/fireworks/models/deepseek-v3-0324","choices":[{"index":0,"delta":{},"finish_reason":"length"}],"usage":{"prompt_tokens":20,"total_tokens":120,"completion_tokens":100}}

data: [DONE]

Parameter

Type

Required

Description

messages

array

Yes

List of messages comprising the conversation

model

string

Yes

Model name (must be prefixed with hf:). See supported Models.

frequency_penalty

number

Penalty for token frequency (-2.0 to 2.0). Reduces repetition.

logit_bias

object

Modify token likelihood. Maps token IDs to bias values (-100 to 100).

logprobs

boolean/number

Return log probabilities of output tokens

max_completion_tokens

number

Maximum tokens for completion (including reasoning tokens)

max_tokens

number

Maximum number of tokens to generate

min_p

number

Minimum probability for nucleus sampling

n

number

Number of completion choices to generate (default: 1)

parallel_tool_calls

boolean

Enable parallel function calling during tool use

presence_penalty

number

Penalty for token presence (-2.0 to 2.0). Encourages new topics.

reasoning_effort

string

Control reasoning effort for thinking models: low, medium, or high

response_format

object

Specify output format (JSON schema, JSON object, etc.)

seed

number

Random seed for deterministic sampling

stop

string/array

Stop sequence(s)

stream

boolean

Stream response using server-sent events

stream_options

object

Options for streaming (when stream: true)

temperature

number

Sampling randomness (0.0-2.0). Higher = more random.

tool_choice

string/object

Control tool calling: auto, none, or specific tool

tools

array

List of tools the model may call

top_k

number

Limit sampling to top K tokens

top_logprobs

number

Number of top token probabilities to return

top_p

number

Nucleus sampling threshold (0.0-1.0)

user

string

Unique identifier representing your end-user

function_call

string/object

(Deprecated) Control function calling: auto, none, or specific function

functions

array

(Deprecated) List of functions the model may call

Parameter

Type

Required

Description

role

string

Yes

Role: system, user, assistant, tool, or function

content

string/array

Yes

Message content (text string or array of content parts)

name

string

Name of the participant (for user/system messages)

tool_calls

array

Tool calls to be executed (for assistant messages)

tool_call_id

string

ID of the tool call (for tool messages)

reasoning_content

string

Reasoning content for thinking models (for assistant messages)

function_call

object

(Deprecated) Function call to be executed (for assistant messages)

import openai client = openai.OpenAI( api_key="SYNTHETIC_API_KEY", base_url="https://api.synthetic.new/openai/v1" ) completion = client.chat.completions.create( model="hf:deepseek-ai/DeepSeek-V3-0324", messages=[ {"role": "system", "content": "You are a helpful assistant."}, {"role": "user", "content": "What is the capital of France?"} ], max_tokens=100, temperature=0.7 ) print(completion.choices[0].message.content)

{ "id": "296a8fcb-181d-46d4-ad9b-b380befa9a07", "object": "chat.completion", "created": 1757644754, "model": "accounts/fireworks/models/deepseek-v3-0324", "choices": [ { "index": 0, "message": { "role": "assistant", "content": "The capital of France is **Paris**. \n\nParis is known for its rich history, iconic landmarks like the Eiffel Tower and the Louvre Museum, and its cultural significance as a global hub for art, fashion, and cuisine. \n\nLet me know if you'd like more details!" }, "finish_reason": "stop" } ], "usage": { "prompt_tokens": 16, "total_tokens": 74, "completion_tokens": 58 } }

completion = client.chat.completions.create( model="hf:deepseek-ai/DeepSeek-V3-0324", messages=[ {"role": "system", "content": "You are a helpful assistant."}, {"role": "user", "content": "Tell me a short story."} ], max_tokens=100, stream=True ) for chunk in completion: if chunk.choices[0].delta.content: print(chunk.choices[0].delta.content, end='', flush=True)

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1757644754,"model":"accounts/fireworks/models/deepseek-v3-0324","choices":[{"index":0,"delta":{"role":"assistant","content":"Once"},"finish_reason":null}]} data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1757644754,"model":"accounts/fireworks/models/deepseek-v3-0324","choices":[{"index":0,"delta":{"content":" upon"},"finish_reason":null}]} data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1757644754,"model":"accounts/fireworks/models/deepseek-v3-0324","choices":[{"index":0,"delta":{"content":" a"},"finish_reason":null}]} data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1757644754,"model":"accounts/fireworks/models/deepseek-v3-0324","choices":[{"index":0,"delta":{},"finish_reason":"length"}],"usage":{"prompt_tokens":20,"total_tokens":120,"completion_tokens":100}} data: [DONE]