@@ -6,6 +6,7 @@ from ml_model.services.deepl import Deepl
 from ml_model.services.epicphotogasm import Epicphotogasm
 from ml_model.services.djourney import Djourney
 from ml_model.services.flux import Flux
+from ml_model.services.granite import Granite
 from ml_model.services.iconic import Iconic
 from ml_model.services.kandinsky import Kandinsky
 from ml_model.services.llama import Llama
@@ -0,0 +1,120 @@
+import time
+
+import requests
+from datetime import timedelta
+from decimal import Decimal
+
+from django.conf import settings
+
+from messages.models import Message, BaseStore
+from ml_model.exceptions.external_api import ExternalAPIException
+from ml_model.models import ModelCategory, ModelInput, ModelParameter
+from ml_model.services.base import SimpleService
+
+
+class Granite(SimpleService):
+    """
+    Granite-3.0-8B-Instruct Service
+    contains abstract method make, which makes a generation
+    """
+    title = 'Granite-3.0-8B-Instruct'
+    description = 'Нейросеть, способная генерировать качественный текст из вашего промпта'
+    category = ModelCategory(title='Чат-боты', slug='chat-bots')
+    versions = []
+    inputs = [ModelInput(type=ModelInput.TypeChoices.TEXT, required=True)]
+    parameters = [
+        ModelParameter(
+            name='Системный промпт',
+            key='system_prompt',
+            type=ModelParameter.TypeChoices.STR,
+            hidden=True
+        ),
+        ModelParameter(
+            name='Лучший процент',
+            key='top_p',
+            type=ModelParameter.TypeChoices.FLOATRANGE,
+            values={'start': 1.0, 'end': 2.0, 'step': 0.1, 'default': 0.9},
+        ),
+        ModelParameter(
+            name='Температура',
+            key='temperature',
+            type=ModelParameter.TypeChoices.FLOATRANGE,
+            values={'start': 0.0, 'end': 1.0, 'step': 0.1, 'default': 0.6},
+        ),
+    ]
+
+    TOKEN_PAYMENT_RULES = {
+        'granite-input': Decimal('27.5'), # 1M tokens
+        'granite-output': Decimal('137.5') # 1M tokens
+    }
+
+    def __init__(self, store: BaseStore) -> None:
+        super().__init__(store)
+        self.urls = {
+            'generate': 'https://api.replicate.com/v1/models/ibm-granite/granite-3.0-8b-instruct/',
+            'get': 'https://api.replicate.com/v1/predictions/',
+        }
+
+    def _call_api(self, payload: dict) -> list:
+        headers = {
+            'Authorization': f'Bearer {settings.REPLICATE_API_KEY}',
+            'Prefer': 'wait',
+        }
+        data = {'input': payload}
+        response = requests.post(
+            url=f'{self.urls['generate']}predictions',
+            headers=headers,
+            json=data,
+        )
+        if response.status_code != 201:
+            raise Exception(response.json())
+        result = requests.get(
+            url=f'{self.urls['get']}{response.json().get('id')}',
+            headers=headers
+        )
+        while result.json()['status'] not in ('succeeded', 'failed', 'canceled'):
+            result = requests.get(
+                url=f'{self.urls['get']}{response.json().get('id')}',
+                headers=headers
+            )
+        return result.json()
+
+    def calculate_price(self, result: str, input_message: Message) -> Decimal:
+        return Decimal(
+            sum(
+                [self.TOKEN_PAYMENT_RULES['granite-output'] / 1_000_000 * len(result.split(' '))]
+                + [self.TOKEN_PAYMENT_RULES['granite-input'] / 1_000_000 * len(input_message.content.split(' '))]
+            )
+        )
+
+    def save_results(
+            self, result: str, time: timedelta, save: bool = True
+    ) -> list[Message]:
+        msgs: list[Message] = [
+            Message(
+                content_object=self.store,
+                content=result,
+                elapsed_time=time,
+            )
+        ]
+        if save:
+            return Message.objects.bulk_create(msgs)
+        return msgs
+
+    def make(self, input_message: Message, save: bool = True) -> list[Message]:
+        start_time = time.time()
+        callback_data = dict(
+            {
+                'prompt': input_message.content,
+                'system_prompt': f'Use only Russian language',
+                **input_message.info,
+            }
+        )
+        try:
+            result = ''.join(self._call_api(payload=callback_data)['output'])
+        except Exception:
+            raise ExternalAPIException
+        process_time = timedelta(seconds=(time.time() - start_time))
+        self.handle_invoice(input_message.content_object.model, result, input_message)
+        msgs = self.save_results(result, process_time, save)
+        return msgs