Add LLMs ti liteLLM config

backend/open_webui/migrations/versions/e0cd6c95cb25_add_time_saved_in_seconds_column_to_.py

@@ -0,0 +1,38 @@
+"""Add time_saved_in_seconds column to completion table
+
+Revision ID: e0cd6c95cb25
+Revises: 8b04db1441b0
+Create Date: 2025-02-25 08:51:45.692183
+
+"""
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+import open_webui.internal.db
+from sqlalchemy.engine.reflection import Inspector
+
+
+# revision identifiers, used by Alembic.
+revision: str = 'e0cd6c95cb25'
+down_revision: Union[str, None] = '8b04db1441b0'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+def column_exists(table, column):
+    conn = op.get_bind()
+    inspector = Inspector.from_engine(conn)
+    columns = inspector.get_columns(table)
+    return any(c["name"] == column for c in columns)
+
+def upgrade() -> None:
+    # Add time_saved_in_seconds column if it doesn't exist
+    if not column_exists("completion", "time_saved_in_seconds"):
+        op.add_column('completion', sa.Column('time_saved_in_seconds', sa.Float (), nullable=True))
+
+
+def downgrade() -> None:
+    with op.batch_alter_table('completion', schema=None) as batch_op:
+        if column_exists("completion", "time_saved_in_seconds"):
+            batch_op.drop_column('time_saved_in_seconds')
+

litellm-config.yaml

@@ -15,7 +15,6 @@ model_list:
       arena_elo: 1287
       knowledge_cutoff: "October 2023"
       context_window: 128000
-
   - model_name: "GPT 3.5 Turbo"
     litellm_params:
       model: "azure/gpt-35-turbo"
@@ -27,5 +26,69 @@ model_list:
       arena_elo: 1116
       knowledge_cutoff: "September 2021"
       context_window: 16000
+  - model_name: 'Claude 3 Haiku'
+    litellm_params:
+      model: 'vertex_ai/claude-3-haiku@20240307'
+      vertex_project: 'testllm-94944'
+      vertex_location: 'europe-west4'
+    model_info:
+      description: 'Lightning-fast responses for simple queries.'
+      arena_elo: 1178
+      knowledge_cutoff: 'August 2023'
+      context_window: 200_000
+  - model_name: 'Claude 3.5 Sonnet'
+    litellm_params:
+      model: 'vertex_ai/claude-3-5-sonnet@20240620'
+      vertex_project: 'testllm-94944'
+      vertex_location: 'europe-west1'
+    model_info:
+      description: "Anthropic's latest model."
+      arena_elo: 1271
+      knowledge_cutoff: 'August 2023'
+      context_window: 200_000
+  - model_name: 'Gemini 1.5 Pro'
+    litellm_params:
+      model: 'gemini-1.5-pro'
+      vertex_project: 'testllm-94944'
+      vertex_location: 'europe-west3'
+    model_info:
+      description: "Google's leading model with a very large context window."
+      arena_elo: 1265
+      knowledge_cutoff: 'November 2023'
+      context_window: 2_000_000
+  - model_name: 'Gemini 1.5 Flash'
+    litellm_params:
+      model: 'gemini-1.5-flash-001'
+      vertex_project: 'testllm-94944'
+      vertex_location: 'europe-west3'
+    model_info:
+      description: 'Optimized for speed and efficiency.'
+      arena_elo: 1231
+      knowledge_cutoff: 'November 2023'
+      context_window: 1_000_000
+  - model_name: 'Mistral Medium'
+    litellm_params:
+      model: mistral/mistral-medium
+    model_info:
+      description: 'Good balance between performance and capacity.'
+      arena_elo: 1148
+      knowledge_cutoff: 'Unknown'
+      context_window: 32_000
+  - model_name: 'Mistral Large'
+    litellm_params:
+      model: mistral/mistral-large-latest
+    model_info:
+      description: 'Ideal for specialized and complex text and code generation.'
+      arena_elo: 1156
+      knowledge_cutoff: 'Unknown'
+      context_window: 32_000
+  - model_name: 'Perplexity'
+    litellm_params:
+      model: 'perplexity/llama-3-sonar-large-32k-online'
+    model_info:
+      max_input_tokens: '32000'
+      input_cost_per_token: 0.000001
+      output_cost_per_token: 0.000001
+      description: 'Optimal for search queries and research.'
 
 general_settings: {}