ASI:One Models

ASI:One offers a family of three models that share the same API surface and capability set. Pick asi1 when you want one model that adapts to the request. Pick asi1-ultra when you need maximum reasoning depth and the largest tool budget. Pick asi1-mini when you want the fastest response and the lightest cost profile.

One family, three tradeoffs

asi1 is the adaptive default and a strong starting point for most workloads. Switch to asi1-ultra for deep reasoning and long agentic runs, or to asi1-mini when latency and cost are the constraint.

asi1

asi1-ultra

asi1-mini

$ curl -X POST https://api.asi1.ai/v1/chat/completions \
>   -H "Content-Type: application/json" \
>   -H "Authorization: Bearer $ASI_ONE_API_KEY" \
>   -d '{
>     "model": "asi1",
>     "messages": [
>       {"role": "user", "content": "Help me plan a trip to Tokyo"}
>     ]
>   }'

Pick your model

asi1

The adaptive default

Balances reasoning depth, speed, and cost. Activates the right capabilities based on your prompt. Start here.

Best for: general agent workflows, chat, tool calling, mixed workloads.

asi1-ultra

Deepest reasoning, largest tool budget

Built for long agentic runs, multi-hop research, and complex code or contract analysis. Up to 500 tool calls per turn.

Best for: long-running tasks, deep research, code audits, agent orchestration.

asi1-mini

Fastest, lightest, low-latency

Optimized for low-latency interactive workloads. Same context window as the rest of the family with a smaller cost and response footprint.

Best for: real-time chat, classification, autocomplete, simple tool calls, voice.

For a side-by-side comparison and a decision tree, see Model Selection.

Shared capabilities

Every model in the family supports the same capability set. The differences are in reasoning depth, tool budget, response size, and latency, not in what the models can do.

Agentic Reasoning

Discovers and orchestrates AI agents from the Agentverse marketplace. Activates when a request benefits from specialized agents - booking, research, scheduling, and more. asi1-ultra sustains the longest agent runs.

Extended Reasoning

Multi-step analysis with intermediate validation. asi1-ultra reaches the deepest, useful for code review, contract analysis, research synthesis, and strategic planning.

Fast Inference

Low-latency responses for real-time applications, chatbots, voice assistants, and interactive experiences. asi1-mini is purpose-built for this profile.

Tool Calling

Define your functions and the model decides when and how to call them. asi1-ultra supports up to 500 tool calls per turn for long agentic flows.

Visualization

Generates charts, graphs, and visual representations of data. Transforms datasets into clear, actionable insights.

Web3 Native

Built for blockchain and decentralized applications. Understands smart contracts, tokenomics, and on-chain interactions.

Specifications at a glance

	`asi1`	`asi1-ultra`	`asi1-mini`
Position	Adaptive default	Deepest reasoning	Fastest response
Context Window	200,000 tokens	200,000 tokens	200,000 tokens
Max Tool Calls per Turn	Standard	Up to 500	Standard
Streaming	Supported	Supported	Supported
OpenAI Compatibility	Full SDK	Full SDK	Full SDK
Chat Completions API	`/v1/chat/completions`	`/v1/chat/completions`	`/v1/chat/completions`
Responses API	`/v1/responses`	`/v1/responses`	`/v1/responses`
Rate Limits	See API dashboard	See API dashboard	See API dashboard

The same API key works across all three models. Switch models by changing the model field in the request body.

Use cases

Common applications by model:

Use Case	Recommended Model	Why
Customer Support	`asi1`	Adapts between fast triage and longer escalation paths
Research and Analysis	`asi1-ultra`	Sustains multi-hop retrieval and synthesis with citations
Code Generation	`asi1`	Handles routine coding tasks with the right depth
Code Review and Audit	`asi1-ultra`	Deepest reasoning for catching subtle issues
Content Creation	`asi1`	General drafting and editing
Real-time Chat	`asi1-mini`	Low latency, short turn-taking
Classification and Routing	`asi1-mini`	Cheap, fast, deterministic
Voice Assistants	`asi1-mini`	Speed-critical, short outputs
Long-running Agent Tasks	`asi1-ultra`	Largest tool budget, deepest plans
Data Visualization	`asi1`	Generates charts and visual explanations
Web3 Applications	`asi1`	Smart contract analysis with adaptive depth

Trust and safety

Feature	Description
Transparent Reasoning	View the model’s thought process with reasoning traces
Reduced Hallucination	Confidence thresholds and cross-checks keep outputs grounded
Safety Filters	Built-in guardrails screen for harmful content
Audit Support	Hashed reasoning steps available for compliance logging

Getting started

Developer Quickstart →

Make your first API request in minutes

Model Selection →

Side-by-side comparison and a decision tree

asi1-ultra →

Deepest reasoning, longest agentic runs

asi1-mini →

Fastest response, lowest latency

One family, three tradeoffs

asi1

asi1-ultra

asi1-mini

$ curl -X POST https://api.asi1.ai/v1/chat/completions \
>   -H "Content-Type: application/json" \
>   -H "Authorization: Bearer $ASI_ONE_API_KEY" \
>   -d '{
>     "model": "asi1",
>     "messages": [
>       {"role": "user", "content": "Help me plan a trip to Tokyo"}
>     ]
>   }'

Pick your model

asi1

The adaptive default

Balances reasoning depth, speed, and cost. Activates the right capabilities based on your prompt. Start here.

Best for: general agent workflows, chat, tool calling, mixed workloads.

asi1-ultra

Deepest reasoning, largest tool budget

Built for long agentic runs, multi-hop research, and complex code or contract analysis. Up to 500 tool calls per turn.

Best for: long-running tasks, deep research, code audits, agent orchestration.

asi1-mini

Fastest, lightest, low-latency

Optimized for low-latency interactive workloads. Same context window as the rest of the family with a smaller cost and response footprint.

Best for: real-time chat, classification, autocomplete, simple tool calls, voice.

For a side-by-side comparison and a decision tree, see Model Selection.

Shared capabilities

Every model in the family supports the same capability set. The differences are in reasoning depth, tool budget, response size, and latency, not in what the models can do.

Agentic Reasoning

Extended Reasoning

Multi-step analysis with intermediate validation. asi1-ultra reaches the deepest, useful for code review, contract analysis, research synthesis, and strategic planning.

Fast Inference

Low-latency responses for real-time applications, chatbots, voice assistants, and interactive experiences. asi1-mini is purpose-built for this profile.

Tool Calling

Define your functions and the model decides when and how to call them. asi1-ultra supports up to 500 tool calls per turn for long agentic flows.

Visualization

Generates charts, graphs, and visual representations of data. Transforms datasets into clear, actionable insights.

Web3 Native

Built for blockchain and decentralized applications. Understands smart contracts, tokenomics, and on-chain interactions.

Specifications at a glance

	`asi1`	`asi1-ultra`	`asi1-mini`
Position	Adaptive default	Deepest reasoning	Fastest response
Context Window	200,000 tokens	200,000 tokens	200,000 tokens
Max Tool Calls per Turn	Standard	Up to 500	Standard
Streaming	Supported	Supported	Supported
OpenAI Compatibility	Full SDK	Full SDK	Full SDK
Chat Completions API	`/v1/chat/completions`	`/v1/chat/completions`	`/v1/chat/completions`
Responses API	`/v1/responses`	`/v1/responses`	`/v1/responses`
Rate Limits	See API dashboard	See API dashboard	See API dashboard

The same API key works across all three models. Switch models by changing the model field in the request body.

Use cases

Common applications by model:

Use Case	Recommended Model	Why
Customer Support	`asi1`	Adapts between fast triage and longer escalation paths
Research and Analysis	`asi1-ultra`	Sustains multi-hop retrieval and synthesis with citations
Code Generation	`asi1`	Handles routine coding tasks with the right depth
Code Review and Audit	`asi1-ultra`	Deepest reasoning for catching subtle issues
Content Creation	`asi1`	General drafting and editing
Real-time Chat	`asi1-mini`	Low latency, short turn-taking
Classification and Routing	`asi1-mini`	Cheap, fast, deterministic
Voice Assistants	`asi1-mini`	Speed-critical, short outputs
Long-running Agent Tasks	`asi1-ultra`	Largest tool budget, deepest plans
Data Visualization	`asi1`	Generates charts and visual explanations
Web3 Applications	`asi1`	Smart contract analysis with adaptive depth

Trust and safety

Feature	Description
Transparent Reasoning	View the model’s thought process with reasoning traces
Reduced Hallucination	Confidence thresholds and cross-checks keep outputs grounded
Safety Filters	Built-in guardrails screen for harmful content
Audit Support	Hashed reasoning steps available for compliance logging

Getting started

Developer Quickstart →

Make your first API request in minutes

Model Selection →

Side-by-side comparison and a decision tree

asi1-ultra →

Deepest reasoning, longest agentic runs

asi1-mini →

Fastest response, lowest latency

$	curl -X POST https://api.asi1.ai/v1/chat/completions \
>	-H "Content-Type: application/json" \
>	-H "Authorization: Bearer $ASI_ONE_API_KEY" \
>	-d '{
>	"model": "asi1",
>	"messages": [
>	{"role": "user", "content": "Help me plan a trip to Tokyo"}
>	]
>	}'