Skip to content
mistral.rs

OpenAI-compatible chat completions endpoint handler.

POST
/v1/chat/completions
 
curl --request POST \
  --url https://example.com/v1/chat/completions \
  --header 'Content-Type: application/json' \
  --data '{ "adapter": { "generation": "5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a5a" }, "agent_permission": null, "code_execution_permission": null, "dry_allowed_length": null, "dry_base": null, "dry_multiplier": null, "dry_sequence_breakers": null, "enable_shell": false, "enable_thinking": null, "files": [ "example" ], "frequency_penalty": null, "grammar": { "type": "regex", "value": "example" }, "logit_bias": null, "logprobs": false, "max_tokens": 256, "max_tool_rounds": null, "messages": [ { "content": "example", "name": "example", "role": "example", "tool_call_id": "example", "tool_calls": [ { "function": { "arguments": "example", "name": "example" }, "id": "example", "type": "function" } ] } ], "min_p": null, "model": "mistral", "n": 1, "presence_penalty": null, "reasoning_effort": null, "repetition_penalty": null, "response_format": { "type": "text" }, "session_id": "example", "stop": "example", "stream": true, "temperature": 0.7, "tool_choice": "none", "tools": null, "top_k": null, "top_logprobs": null, "top_p": null, "truncate_sequence": null, "web_search_options": { "external_web_access": true, "extract_description": "example", "filters": { "allowed_domains": [ "example" ], "blocked_domains": [ "example" ] }, "image_settings": { "caption": true, "max_results": 1 }, "return_token_budget": "default", "search_content_types": [ "text" ], "search_context_size": "low", "search_description": "example", "user_location": { "approximate": { "city": "example", "country": "example", "region": "example", "timezone": "example" }, "type": "approximate" } } }'

Request Body required

Section titled “Request Body required ”
Media type application/json

Chat completion request following OpenAI’s specification

object
adapter
One of:
null
agent_permission

Permission policy for agentic tools.

string | null
Example 
null
code_execution_permission

Permission policy for code execution.

string | null
Example 
null
dry_allowed_length

Longest repeated sequence DRY leaves unpenalized.

integer | null
Example 
null
dry_base

Base for DRY’s exponential penalty growth.

number | null format: float
Example 
null
dry_multiplier

DRY repetition penalty multiplier; 0 disables DRY.

number | null format: float
Example 
null
dry_sequence_breakers

Sequences that reset DRY repetition matching.

Array<string> | null
Example 
null
enable_shell

Enable the built-in shell execution tool.

boolean
Example 
false
enable_thinking

Toggle thinking output for models that support it.

boolean | null
Example 
null
files

Required output files. The runtime asks the model to produce them and surfaces a File (or error placeholder) for each.

Array | null
frequency_penalty

Penalize tokens by how often they have appeared so far; positive values reduce repetition.

number | null format: float
Example 
null
grammar
One of:
null
logit_bias

Bias added to the logits of these token IDs before sampling.

object | null
Example 
null
logprobs

Return log probabilities of the output tokens.

boolean
Example 
false
max_tokens

Maximum number of tokens to generate.

integer | null
Example 
256
max_tool_rounds

Maximum number of tool-call rounds the server will auto-execute.

integer | null
Example 
null
messages
required
One of:
Array<object>

Represents a single message in a conversation

Examples

use either::Either;
use mistralrs_server_core::openai::{Message, MessageContent};

// User message
let user_msg = Message {
    content: Some(MessageContent(Either::Left("What's 2+2?".to_string()))),
    role: "user".to_string(),
    name: None,
    tool_calls: None,
};

// System message
let system_msg = Message {
    content: Some(MessageContent(Either::Left("You are a helpful assistant.".to_string()))),
    role: "system".to_string(),
    name: None,
    tool_calls: None,
};
object
content
One of:
null
name

Optional participant name for this message

string | null
role
required

The role of the message sender (“user”, “assistant”, “system”, “tool”, etc.)

string
tool_call_id

Tool call ID this message is responding to (for tool messages)

string | null
tool_calls

Optional list of tool calls (for assistant messages)

Array<object> | null

Represents a tool call made by the assistant

This structure wraps a function call with its type information.

object
function
required

The function call details

object
arguments
required

The function arguments (JSON string)

string
name
required

The name of the function to call

string
id

Unique identifier for this tool call

string | null
type
required

The type of tool being called

string
Allowed values: function
min_p

Drop tokens below this fraction of the top token’s probability.

number | null format: double
Example 
null
model

Model ID; “default” targets the only loaded model.

string
Example 
mistral
n

How many choices to generate.

integer
Example 
1
presence_penalty

Penalize tokens that have already appeared; positive values push toward new topics.

number | null format: float
Example 
null
reasoning_effort

Reasoning effort level for Harmony-format models (GPT-OSS). Controls the depth of reasoning/analysis: “low”, “medium”, or “high”.

string | null
Example 
null
repetition_penalty

Multiplicative repetition penalty; 1.0 disables it.

number | null format: float
Example 
null
response_format
One of:
null
session_id

Persistent agentic state. If None, a new session is created and the ID is returned in the response.

string | null
stop
One of:
null
stream

Stream the response as server-sent events.

boolean | null
Example 
true
temperature

Sampling temperature; higher values increase randomness.

number | null format: double
Example 
0.7
tool_choice
One of:
null
tools

Tools the model may call.

Array | null
One of:

Tool definition

object
function
required

Function definition for a tool

object
description
string | null
name
required
string
parameters
object | null
strict

When true, the tool’s parameters JSON schema is enforced on the generated arguments via constrained decoding (llguidance).

boolean | null
type
required

Type of tool

string
Allowed values: function
Example 
null
top_k

Sample only from the k most likely tokens.

integer | null
Example 
null
top_logprobs

Number of most likely tokens to return per position; requires logprobs.

integer | null
Example 
null
top_p

Nucleus sampling: only tokens within the top cumulative probability mass are considered.

number | null format: double
Example 
null
truncate_sequence

Truncate inputs that exceed the model’s context length instead of erroring.

boolean | null
Example 
null
web_search_options
One of:
null

Responses

Section titled “ Responses ”

200

Section titled “200 ”

Chat completion JSON or server-sent event chunks

object
adapter_generation
string | null
agentic_tool_calls
Array | null
choices
required
Array<object>
object
finish_reason
required
string
index
required
integer
logprobs
message
required
object
content
string | null
reasoning_content
string | null
role
required
string
tool_calls
Array | null
created
required
integer format: int64
files
Array | null
id
required
string
model
required
string
object
required
string
session_id
string | null
system_fingerprint
required
string
usage
required
object
avg_compl_tok_per_sec
required
number format: float
avg_prompt_tok_per_sec
required
number format: float
avg_tok_per_sec
required
number format: float
completion_tokens
required
integer
prompt_tokens
required
integer
total_completion_time_sec
required
number format: float
total_prompt_time_sec
required
number format: float
total_time_sec
required
number format: float
total_tokens
required
integer
Example generated
{
  "adapter_generation": "example",
  "agentic_tool_calls": [
    "example"
  ],
  "choices": [
    {
      "finish_reason": "example",
      "index": 1,
      "logprobs": "example",
      "message": {
        "content": "example",
        "reasoning_content": "example",
        "role": "example",
        "tool_calls": [
          "example"
        ]
      }
    }
  ],
  "created": 1,
  "files": [
    "example"
  ],
  "id": "example",
  "model": "example",
  "object": "example",
  "session_id": "example",
  "system_fingerprint": "example",
  "usage": {
    "avg_compl_tok_per_sec": 1,
    "avg_prompt_tok_per_sec": 1,
    "avg_tok_per_sec": 1,
    "completion_tokens": 1,
    "prompt_tokens": 1,
    "total_completion_time_sec": 1,
    "total_prompt_time_sec": 1,
    "total_time_sec": 1,
    "total_tokens": 1
  }
}