Model Differences
Models vary in context limits, capabilities, and supported request formats.
Conversational AI
ConversationalAI that TurnsKnowledge intoAction
Access supported third-party models for reasoning, research, and writing through one interface, with model-specific pricing shown before use.
Model accessSupported third-party providers
Input + output
Problem
Managing Multiple Models Adds Complexity
Model capabilities, provider requirements, and token prices differ across the market. MARIA AI brings these choices into one interface.
Provider Requirements
Each provider may apply different terms, availability limits, and data practices.
Cost Visibility
Input and output prices differ by model and can change as upstream providers update rates.
Solution
Unified Access for Research, Reasoning, and Writing
Use supported third-party models through one interface, with model selection and published usage pricing.
Docs
MARIA AIThird-party Model Access Platform
Action
Research Workflows
Use supported models to summarize and organize documents within your review process.
AI Assistants
Build assistants with model selection that matches your organization's requirements.
API-Compatible Integration
Send supported requests through one interface and review usage by model.
Model Library
Third-Party Models with Published Rates
Unified Model Access
Use one consistent interface for supported third-party models.
Usage-Based Billing
Input and output tokens are charged at published per-model rates.
Provider Transparency
Availability, limits, and data handling depend on the selected provider.
Pricing
Clear, Usage-Based Model Pricing
Pay only for model usage. Input and output tokens are billed separately at the published rate for your selected model.
Pay as you go
$0monthly minimum
- Access supported models through the MARIA AI API
- Model selection and usage reporting
- Email support
Model usage is billed separately.
Open-weight model usage
Current starting rates in USD per 1M tokens.
Nemotron 3.5 LightningNVIDIA
Input / 1M$0.08
Output / 1M$0.20
Qwen3.8 27BQwen
Input / 1M$0.35
Output / 1M$2.75
DeepSeek V4 Pro 0813DeepSeek
Input / 1M$0.66
Output / 1M$1.98
Qwen3.8 2.4T A95BQwen
Input / 1M$2.00
Output / 1M$6.00
Rates shown as of August 24, 2026. Prices may change; the current rate is confirmed before purchase.
MARIA AI provides paid aggregation and routing access to third-party models. We do not own or develop the listed models. Provider terms and acceptable use policies apply.
For Your Team
Choose Models with Clear Terms and Costs
Choose supported models, review provider terms, and plan usage with published pricing.
Developers and Product teams
- Build assistants with supported third-party models
- Switch models through a consistent API interface
- Track input and output usage by project
Platform and AI Infra teams
- Compare supported models by capability and token price
- Route requests through one consistent API interface
- Review provider requirements before production deployment
Innovation and Strategy Leaders
- Compare model costs before selecting a provider
- Standardize model access across approved workflows
- Plan usage with visible token-based rates