Complétion de conversation Grok (tous modèles, paramètres complets)

curl --request POST \
  --url https://direct.evolink.ai/v1/chat/completions \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "grok-4.5",
  "messages": [
    {
      "role": "system",
      "content": "You are a concise assistant."
    },
    {
      "role": "user",
      "content": "Explain prompt caching in one sentence."
    }
  ],
  "stream": false,
  "max_tokens": 1024,
  "temperature": 0.7,
  "top_p": 0.95,
  "tools": [
    {
      "type": "function",
      "function": {}
    }
  ]
}
'

import requests

url = "https://direct.evolink.ai/v1/chat/completions"

payload = {
    "model": "grok-4.5",
    "messages": [
        {
            "role": "system",
            "content": "You are a concise assistant."
        },
        {
            "role": "user",
            "content": "Explain prompt caching in one sentence."
        }
    ],
    "stream": False,
    "max_tokens": 1024,
    "temperature": 0.7,
    "top_p": 0.95,
    "tools": [
        {
            "type": "function",
            "function": {}
        }
    ]
}
headers = {
    "Authorization": "Bearer <token>",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {Authorization: 'Bearer <token>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    model: 'grok-4.5',
    messages: [
      {role: 'system', content: 'You are a concise assistant.'},
      {role: 'user', content: 'Explain prompt caching in one sentence.'}
    ],
    stream: false,
    max_tokens: 1024,
    temperature: 0.7,
    top_p: 0.95,
    tools: [{type: 'function', function: {}}]
  })
};

fetch('https://direct.evolink.ai/v1/chat/completions', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://direct.evolink.ai/v1/chat/completions",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'model' => 'grok-4.5',
    'messages' => [
        [
                'role' => 'system',
                'content' => 'You are a concise assistant.'
        ],
        [
                'role' => 'user',
                'content' => 'Explain prompt caching in one sentence.'
        ]
    ],
    'stream' => false,
    'max_tokens' => 1024,
    'temperature' => 0.7,
    'top_p' => 0.95,
    'tools' => [
        [
                'type' => 'function',
                'function' => [
                                
                ]
        ]
    ]
  ]),
  CURLOPT_HTTPHEADER => [
    "Authorization: Bearer <token>",
    "Content-Type: application/json"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://direct.evolink.ai/v1/chat/completions"

	payload := strings.NewReader("{\n  \"model\": \"grok-4.5\",\n  \"messages\": [\n    {\n      \"role\": \"system\",\n      \"content\": \"You are a concise assistant.\"\n    },\n    {\n      \"role\": \"user\",\n      \"content\": \"Explain prompt caching in one sentence.\"\n    }\n  ],\n  \"stream\": false,\n  \"max_tokens\": 1024,\n  \"temperature\": 0.7,\n  \"top_p\": 0.95,\n  \"tools\": [\n    {\n      \"type\": \"function\",\n      \"function\": {}\n    }\n  ]\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Authorization", "Bearer <token>")
	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://direct.evolink.ai/v1/chat/completions")
  .header("Authorization", "Bearer <token>")
  .header("Content-Type", "application/json")
  .body("{\n  \"model\": \"grok-4.5\",\n  \"messages\": [\n    {\n      \"role\": \"system\",\n      \"content\": \"You are a concise assistant.\"\n    },\n    {\n      \"role\": \"user\",\n      \"content\": \"Explain prompt caching in one sentence.\"\n    }\n  ],\n  \"stream\": false,\n  \"max_tokens\": 1024,\n  \"temperature\": 0.7,\n  \"top_p\": 0.95,\n  \"tools\": [\n    {\n      \"type\": \"function\",\n      \"function\": {}\n    }\n  ]\n}")
  .asString();

require 'uri'
require 'net/http'

url = URI("https://direct.evolink.ai/v1/chat/completions")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Authorization"] = 'Bearer <token>'
request["Content-Type"] = 'application/json'
request.body = "{\n  \"model\": \"grok-4.5\",\n  \"messages\": [\n    {\n      \"role\": \"system\",\n      \"content\": \"You are a concise assistant.\"\n    },\n    {\n      \"role\": \"user\",\n      \"content\": \"Explain prompt caching in one sentence.\"\n    }\n  ],\n  \"stream\": false,\n  \"max_tokens\": 1024,\n  \"temperature\": 0.7,\n  \"top_p\": 0.95,\n  \"tools\": [\n    {\n      \"type\": \"function\",\n      \"function\": {}\n    }\n  ]\n}"

response = http.request(request)
puts response.read_body

{
  "id": "chatcmpl-20260812164515123456789AbCdEfGh",
  "model": "grok-4.5",
  "object": "chat.completion",
  "created": 1786538000,
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Prompt caching reuses previously processed prompt prefixes so repeated context is billed at a lower rate.",
        "tool_calls": [
          {}
        ]
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 504,
    "completion_tokens": 2,
    "total_tokens": 526,
    "prompt_tokens_details": {
      "cached_tokens": 0
    }
  }
}

{
  "error": {
    "code": 400,
    "message": "Invalid request parameters",
    "type": "invalid_request_error"
  }
}

{
  "error": {
    "code": 401,
    "message": "Invalid or expired token",
    "type": "authentication_error"
  }
}

{
  "error": {
    "code": 402,
    "message": "Insufficient quota",
    "type": "insufficient_quota_error",
    "fallback_suggestion": "https://evolink.ai/dashboard/billing"
  }
}

{
  "error": {
    "code": 429,
    "message": "Rate limit exceeded",
    "type": "rate_limit_error",
    "fallback_suggestion": "retry after 60 seconds"
  }
}

{
  "error": {
    "code": 500,
    "message": "Internal server error",
    "type": "internal_server_error",
    "fallback_suggestion": "try again later"
  }
}

{
  "error": {
    "code": 503,
    "message": "Service temporarily unavailable",
    "type": "service_unavailable_error",
    "fallback_suggestion": "retry after 30 seconds"
  }
}

Chat Completions API

API tous modèles Grok - Référence complète Chat Completions

Point de terminaison Chat Completions compatible OpenAI pour les modèles de texte xAI Grok ; sélection du modèle via le paramètre model (toutes les valeurs figurent dans le tableau du paramètre model)
grok-4.5 : fenêtre de contexte de 500K tokens ; à partir de 200K tokens dans le prompt, tous les types de tokens sont facturés au double du tarif
La mise en cache du prompt est automatique : les tokens servis depuis le cache sont facturés au tarif d’entrée en cache, moins élevé
Modes synchrone et streaming (SSE)
Les appels d’outils function classiques sont pris en charge ; les outils côté serveur de xAI sont disponibles uniquement via la Responses API

POST

chat

completions

Complétion de conversation Grok (tous modèles, paramètres complets)

curl --request POST \
  --url https://direct.evolink.ai/v1/chat/completions \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "grok-4.5",
  "messages": [
    {
      "role": "system",
      "content": "You are a concise assistant."
    },
    {
      "role": "user",
      "content": "Explain prompt caching in one sentence."
    }
  ],
  "stream": false,
  "max_tokens": 1024,
  "temperature": 0.7,
  "top_p": 0.95,
  "tools": [
    {
      "type": "function",
      "function": {}
    }
  ]
}
'

import requests

url = "https://direct.evolink.ai/v1/chat/completions"

payload = {
    "model": "grok-4.5",
    "messages": [
        {
            "role": "system",
            "content": "You are a concise assistant."
        },
        {
            "role": "user",
            "content": "Explain prompt caching in one sentence."
        }
    ],
    "stream": False,
    "max_tokens": 1024,
    "temperature": 0.7,
    "top_p": 0.95,
    "tools": [
        {
            "type": "function",
            "function": {}
        }
    ]
}
headers = {
    "Authorization": "Bearer <token>",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {Authorization: 'Bearer <token>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    model: 'grok-4.5',
    messages: [
      {role: 'system', content: 'You are a concise assistant.'},
      {role: 'user', content: 'Explain prompt caching in one sentence.'}
    ],
    stream: false,
    max_tokens: 1024,
    temperature: 0.7,
    top_p: 0.95,
    tools: [{type: 'function', function: {}}]
  })
};

fetch('https://direct.evolink.ai/v1/chat/completions', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://direct.evolink.ai/v1/chat/completions",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'model' => 'grok-4.5',
    'messages' => [
        [
                'role' => 'system',
                'content' => 'You are a concise assistant.'
        ],
        [
                'role' => 'user',
                'content' => 'Explain prompt caching in one sentence.'
        ]
    ],
    'stream' => false,
    'max_tokens' => 1024,
    'temperature' => 0.7,
    'top_p' => 0.95,
    'tools' => [
        [
                'type' => 'function',
                'function' => [
                                
                ]
        ]
    ]
  ]),
  CURLOPT_HTTPHEADER => [
    "Authorization: Bearer <token>",
    "Content-Type: application/json"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://direct.evolink.ai/v1/chat/completions"

	payload := strings.NewReader("{\n  \"model\": \"grok-4.5\",\n  \"messages\": [\n    {\n      \"role\": \"system\",\n      \"content\": \"You are a concise assistant.\"\n    },\n    {\n      \"role\": \"user\",\n      \"content\": \"Explain prompt caching in one sentence.\"\n    }\n  ],\n  \"stream\": false,\n  \"max_tokens\": 1024,\n  \"temperature\": 0.7,\n  \"top_p\": 0.95,\n  \"tools\": [\n    {\n      \"type\": \"function\",\n      \"function\": {}\n    }\n  ]\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Authorization", "Bearer <token>")
	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://direct.evolink.ai/v1/chat/completions")
  .header("Authorization", "Bearer <token>")
  .header("Content-Type", "application/json")
  .body("{\n  \"model\": \"grok-4.5\",\n  \"messages\": [\n    {\n      \"role\": \"system\",\n      \"content\": \"You are a concise assistant.\"\n    },\n    {\n      \"role\": \"user\",\n      \"content\": \"Explain prompt caching in one sentence.\"\n    }\n  ],\n  \"stream\": false,\n  \"max_tokens\": 1024,\n  \"temperature\": 0.7,\n  \"top_p\": 0.95,\n  \"tools\": [\n    {\n      \"type\": \"function\",\n      \"function\": {}\n    }\n  ]\n}")
  .asString();

require 'uri'
require 'net/http'

url = URI("https://direct.evolink.ai/v1/chat/completions")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Authorization"] = 'Bearer <token>'
request["Content-Type"] = 'application/json'
request.body = "{\n  \"model\": \"grok-4.5\",\n  \"messages\": [\n    {\n      \"role\": \"system\",\n      \"content\": \"You are a concise assistant.\"\n    },\n    {\n      \"role\": \"user\",\n      \"content\": \"Explain prompt caching in one sentence.\"\n    }\n  ],\n  \"stream\": false,\n  \"max_tokens\": 1024,\n  \"temperature\": 0.7,\n  \"top_p\": 0.95,\n  \"tools\": [\n    {\n      \"type\": \"function\",\n      \"function\": {}\n    }\n  ]\n}"

response = http.request(request)
puts response.read_body

{
  "id": "chatcmpl-20260812164515123456789AbCdEfGh",
  "model": "grok-4.5",
  "object": "chat.completion",
  "created": 1786538000,
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Prompt caching reuses previously processed prompt prefixes so repeated context is billed at a lower rate.",
        "tool_calls": [
          {}
        ]
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 504,
    "completion_tokens": 2,
    "total_tokens": 526,
    "prompt_tokens_details": {
      "cached_tokens": 0
    }
  }
}

{
  "error": {
    "code": 400,
    "message": "Invalid request parameters",
    "type": "invalid_request_error"
  }
}

{
  "error": {
    "code": 401,
    "message": "Invalid or expired token",
    "type": "authentication_error"
  }
}

{
  "error": {
    "code": 402,
    "message": "Insufficient quota",
    "type": "insufficient_quota_error",
    "fallback_suggestion": "https://evolink.ai/dashboard/billing"
  }
}

{
  "error": {
    "code": 429,
    "message": "Rate limit exceeded",
    "type": "rate_limit_error",
    "fallback_suggestion": "retry after 60 seconds"
  }
}

{
  "error": {
    "code": 500,
    "message": "Internal server error",
    "type": "internal_server_error",
    "fallback_suggestion": "try again later"
  }
}

{
  "error": {
    "code": 503,
    "message": "Service temporarily unavailable",
    "type": "service_unavailable_error",
    "fallback_suggestion": "retry after 30 seconds"
  }
}

BaseURL : La BaseURL par défaut est https://direct.evolink.ai, qui offre une meilleure prise en charge des modèles de texte et des connexions persistantes. https://api.evolink.ai est le point d’accès principal pour les services multimodaux et sert d’adresse de secours pour les modèles de texte.

Les outils côté serveur (recherche web, recherche X, exécution de code, recherche de pièces jointes, recherche de collections) sont disponibles uniquement via la Responses API. Le point de terminaison Chat Completions ne prend en charge que les appels d’outils function classiques.

Autorisations

Authorization

string

header

requis

##Toutes les API nécessitent une authentification Bearer Token##

Obtenir une clé API :

Visitez la Page de gestion des clés API pour obtenir votre clé API

Ajouter à l'en-tête de requête :

Authorization: Bearer YOUR_API_KEY

Corps

application/json

model

enum<string>

requis

Modèle à appeler :

ID du modèle	Positionnement
`grok-4.5`	Modèle xAI de raisonnement et d'appel d'outils, fenêtre de contexte de 500K

Options disponibles:

grok-4.5

Exemple:

"grok-4.5"

messages

object[]

requis

Liste des messages de chat. Prend en charge les rôles system, user et assistant.

Minimum array length: 1

Show child attributes

Exemple:

[
  {
    "role": "system",
    "content": "You are a concise assistant."
  },
  {
    "role": "user",
    "content": "Explain prompt caching in one sentence."
  }
]

stream

boolean

défaut:false

Indique s'il faut renvoyer une réponse en streaming (SSE, événements chat.completion.chunk). Par défaut false.

Exemple:

false

max_tokens

integer

Nombre maximal de tokens à générer. Transmis tel quel au modèle.

Exemple:

1024

temperature

number

Température d'échantillonnage (0-2). Des valeurs plus élevées produisent une sortie plus aléatoire.

Exemple:

0.7

top_p

number

Paramètre d'échantillonnage nucleus (0-1).

Exemple:

0.95

tools

object[]

Définitions d'outils function OpenAI classiques (appels de fonctions côté client, sans frais supplémentaires par appel). Les outils côté serveur de xAI sont disponibles uniquement via la Responses API.

Show child attributes

tool_choice

Contrôle la sélection de la fonction : "auto" / "none" / "required", ou un objet imposant une fonction précise.

Options disponibles:

auto,

none,

required

Réponse

Complétion de chat générée avec succès (objet JSON, ou flux SSE d'événements chat.completion.chunk lorsque stream=true)

string

Identifiant unique pour la complétion de chat

Exemple:

"chatcmpl-20260812164515123456789AbCdEfGh"

model

string

Nom du modèle réellement utilisé

Exemple:

"grok-4.5"

object

enum<string>

Type de réponse

Options disponibles:

chat.completion

Exemple:

"chat.completion"

created

integer

Horodatage de création

Exemple:

1786538000

choices

object[]

Liste des choix de complétion de chat

Show child attributes

usage

object

Statistiques d'utilisation des tokens. À partir de 200K tokens dans le prompt, tous les types de tokens (entrée, entrée en cache, sortie) sont facturés au double.

Show child attributes

Série Image

Série Vidéo

Série Audio

Série Texte

Gestion du compte

Gestion des tâches

Gestion des fichiers

API tous modèles Grok - Référence complète Chat Completions

Autorisations

Corps

Réponse