Omni AI
AI model · Language model (LLM) · Text → Text by Omni
The proprietary AI model for trusted, semantic-aware analytics.
Omni's proprietary free AI model for governed data analytics. It provides businesses with a trusted way to query data using natural language, backed by a 400,000-character prompt limit and built-in query verification.
Model specs
The full spec card (context, pricing per token, modalities and more) on AntSpace. Sign up free to see the full spec card.
- Price
- From 0 USD
- Launched
- 2026-11-02
- Languages
- English
Category: Generative AI
- Modalities
- Text to Text
- License
- Proprietary
- Access
- Early access via account manager or Omni support
- Released
- 2026-11-02
- Prompt Limit
- 400,000 characters
- Execution Profiles
- Data-only, Full Agent Session
- Proprietary free AI model for Omni users
- 400,000-character prompt limit
- Built-in query verification for semantic accuracy
- Data-only execution profile for single questions
Overview
Omni AI is Omni's proprietary language model designed specifically for data analytics and business intelligence workflows. It serves as a free model option for organizations on the Omni platform, operating directly on top of a governed semantic layer to ensure trustworthy results. The model supports advanced reasoning features, such as a data-only execution profile for single questions and a verification object that details how generated queries interact with the semantic model. It is currently available in early access for businesses looking to optimize their AI analytics with a model that understands their specific business logic and metrics.
Reasoning and Verification
- Semantic Layer Integration: Omni AI operates directly on top of the Omni semantic layer to ensure data logic and metrics are consistent across all queries.
- Data-Only Execution Profile: The model can answer single data questions using a restricted, query-only toolset for focused analysis.
- Query Verification Object: Omni AI provides facts about how a generated query used the semantic model, including the use of CTEs and custom fields.
- Result Identification: The model identifies the final query's executed result with a resultId to allow for correlation across different sessions.
Access and Limits
- Free Model Access: Omni AI is available as a free model option for organizations using the Omni platform.
- Early Access Program: Users can access the model in early access by contacting their account manager or Omni support.
Use cases
- Data analysts use Omni AI to generate complex SQL queries that automatically adhere to defined semantic metrics.
- Business users ask single data questions using the data-only execution profile to get rapid answers without a full agent session.
- Administrators review verification objects to confirm that AI-generated queries are using the correct joins and custom fields.
Specifications
- Modalities
- Text to Text
- License
- Proprietary
- Access
- Early access via account manager or Omni support
- Released
- 2026-11-02
- Prompt Limit
- 400,000 characters
- Execution Profiles
- Data-only, Full Agent Session
- Verification Features
- Query verification objects, resultId correlation
- Access Method
- Early Access via Account Manager
- Integration
- Omni Semantic Layer
How it works
- Request Early Access
- Enable in AI Settings: Once enabled, the model can be selected within the Omni platform's AI interfaces, such as the Omni Agent or chat panel.
- Query and Verify: Users submit natural language data questions, which the model translates into queries grounded in the organization's semantic layer.
Pricing
- Omni AI (Early Access)
- Free for Omni organizations
FAQ
What is Omni AI?
Omni AI is Omni's proprietary free AI model, currently available in early access for organizations on the platform.
How do I get access to Omni AI?
You can enable Omni AI for your organization by contacting your account manager or Omni support.
How does Omni AI ensure query accuracy?
The model includes a verification object that provides facts about how it used the semantic model, such as whether it stayed on topic or used custom fields.
What are the prompt limits for Omni AI?
Prompts sent to the model are limited to 400,000 characters; drafts over this limit stay in the chat box for shortening.
Added by AntSpace from public sources (docs.omni.co).