> ## Documentation Index
> Fetch the complete documentation index at: https://docs.zerotwo.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Models & Providers

> Overview of all AI models and providers available on ZeroTwo, with capabilities and use cases

## Available Models

ZeroTwo provides access to 50+ premium AI models from the world's leading providers, allowing you to choose the best model for each specific task.

<Note>
  Switch between models mid-conversation to leverage each model's unique strengths. No need to start a new chat or lose context.
</Note>

## Chat Models

### OpenAI

Industry-leading models for general-purpose AI tasks with advanced reasoning capabilities.

<CardGroup cols={2}>
  <Card title="GPT-5 Series" icon="sparkles">
    **Models**: GPT-5, GPT-5-mini, GPT-5-nano

    Next-generation flagship models with enhanced reasoning and multimodal understanding.
  </Card>

  <Card title="GPT-4.1 Series" icon="star">
    **Models**: GPT-4.1, GPT-4.1-mini, GPT-4.1-nano

    Latest GPT-4 generation with improved performance and cost efficiency.
  </Card>

  <Card title="GPT-4o Series" icon="robot">
    **Models**: GPT-4o, GPT-4o-mini

    Optimized models balancing performance and speed for general tasks.
  </Card>

  <Card title="Reasoning Models" icon="brain">
    **Models**: O3, O4-mini

    Specialized models for complex reasoning, mathematics, and logic problems.
  </Card>
</CardGroup>

**Key Capabilities**:

* Advanced reasoning and problem-solving
* Code generation and debugging
* Multimodal understanding (text, images, documents)
* Streaming responses for real-time interaction
* Function calling for tool integration

**Best For**: Code generation, creative writing, complex analysis, general-purpose tasks

***

### Anthropic

Long-context specialists with exceptional document analysis capabilities.

**Available Models**:

* **claude-sonnet-4-20250514**: Latest Sonnet with enhanced capabilities
* **claude-3-5-sonnet-20241022**: Balanced performance and intelligence
* **claude-3-5-haiku-20241022**: Fast, efficient model for quick tasks

**Key Capabilities**:

* Extended context windows (up to 200K tokens)
* Superior document analysis and summarization
* Advanced reasoning with explicit thinking
* Ethical and safe responses
* Excellent instruction following

**Best For**: Long document analysis, research papers, legal documents, detailed technical writing

***

### Google Gemini

Multimodal excellence with real-time web integration.

**Available Models**:

* **gemini-2.5-pro**: Flagship model with comprehensive capabilities
* **gemini-2.5-flash**: Fast, efficient for most tasks
* **gemini-2.0-flash**: Optimized for speed
* **gemini-1.5-pro**: Previous generation, still highly capable

**Key Capabilities**:

* Native multimodal understanding
* Real-time information integration
* Visual analysis and description
* Fast processing speeds
* YouTube and web content analysis

**Best For**: Multimodal tasks, image analysis, video understanding, real-time information needs

***

### Mistral AI

Open-source excellence with specialized knowledge processing.

**Available Models**:

* **mistral-large-latest**: Flagship model for complex tasks
* **mistral-small-latest**: Efficient for common workflows
* **open-mistral-nemo**: Community-driven open model

**Key Capabilities**:

* Efficient processing with lower latency
* Specialized knowledge domains
* Open-source transparency
* Cost-effective performance
* European data sovereignty options

**Best For**: Privacy-conscious applications, cost-effective solutions, specialized domains

***

### DeepSeek

Specialized reasoning with innovative thinking models.

**Available Models**:

* **deepseek-chat**: General-purpose conversational AI
* **deepseek-reasoner**: Specialized for complex reasoning

**Key Capabilities**:

* Advanced reasoning capabilities
* Chain-of-thought processing
* Mathematical problem-solving
* Logical inference
* Research-focused design

**Best For**: Scientific research, mathematical problems, logical reasoning, academic tasks

***

### Groq

Ultra-fast inference with exceptional processing speed.

**Available Models**:

* **llama-3.3-70b-versatile**: High-performance Llama 3
* **moonshotai/kimi-k2-instruct**: Specialized instruction following

**Key Capabilities**:

* Ultra-fast inference speed
* High-throughput processing
* Low latency responses
* Efficient token usage
* Real-time applications

**Best For**: Real-time applications, high-volume processing, latency-critical tasks

***

### X.AI (xAI)

Uncensored models with real-time web access and current information.

**Available Models**:

* **grok-4-0709**: Most advanced Grok model
* **grok-3**: Flagship general model
* **grok-3-fast**: Speed-optimized variant
* **grok-3-mini**: Compact, efficient model
* **grok-3-mini-fast**: Ultra-fast compact model

**Key Capabilities**:

* Uncensored and direct responses
* Real-time web access built-in
* Current events knowledge
* Unfiltered information
* X (Twitter) integration

**Best For**: Current events, unfiltered analysis, real-time information, social media insights

***

### Cohere

Enterprise-focused with strong knowledge processing capabilities.

**Available Models**:

* **command-a-03-2025**: Latest command model
* **command-r7b-12-2024**: Efficient 7B parameter model
* **command-r-plus**: Enhanced command capabilities
* **command-r**: Standard command model

**Key Capabilities**:

* Enterprise-grade reliability
* Knowledge base integration
* Document processing
* Semantic search
* RAG (Retrieval-Augmented Generation)

**Best For**: Enterprise applications, document search, knowledge management, business intelligence

***

### Perplexity

Web search specialists with research-grade capabilities.

**Available Models**:

* **sonar-pro**: Advanced research capabilities
* **sonar**: Standard web-augmented model

**Key Capabilities**:

* Native web search integration
* Real-time information access
* Cited sources and references
* Research-grade accuracy
* Current event awareness

**Best For**: Research, fact-checking, current information, cited content

***

### Qwen (Alibaba)

Multilingual excellence with strong reasoning capabilities.

**Available Models**:

* **qwen-max-latest**: Flagship maximum capability model
* **qwen-flash**: Fast, efficient processing
* **qwen-turbo-latest**: Speed-optimized model
* **qwen3-235b-a22b-instruct-2507**: Large-scale instruction model
* **qwen3-235b-a22b-thinking-2507**: Specialized thinking model

**Key Capabilities**:

* Exceptional multilingual support
* Strong reasoning abilities
* Large-scale model architectures
* Thinking and reflection models
* Chinese language expertise

**Best For**: Multilingual tasks, Chinese language, reasoning problems, large-context needs

***

### OpenRouter

Access to specialized open-source models.

**Available Models**:

* **cognitivecomputations/dolphin-mixtral-8x22b**: Uncensored, helpful assistant

**Key Capabilities**:

* Uncensored responses
* Open-source transparency
* Specialized fine-tuning
* Community-driven improvements
* Flexible usage policies

**Best For**: Uncensored applications, open-source projects, specialized fine-tunes

***

### Liquid

Innovative language models with unique architectures.

**Available Models**:

* **liquid/lfm-7b**: 7B parameter liquid foundation model

**Key Capabilities**:

* Novel architecture designs
* Efficient processing
* Specialized applications
* Research-driven development

**Best For**: Experimental applications, research projects, novel AI approaches

***

### Thesys

Cutting-edge experimental models with enhanced capabilities.

**Available Models**:

* **c1/anthropic/claude-sonnet-4/v-20250815**: Enhanced Claude Sonnet 4 variant

**Key Capabilities**:

* Experimental enhancements
* Cutting-edge performance
* Research-grade quality
* Advanced capabilities

**Best For**: Advanced users, research projects, cutting-edge AI applications

## Image Generation Models

Create stunning images with top-tier AI models.

<CardGroup cols={2}>
  <Card title="gpt-image-1" icon="https://mintcdn.com/zerotwo/6HRSbZHwyN84Av-J/images/icons/providers/openai.svg?fit=max&auto=format&n=6HRSbZHwyN84Av-J&q=85&s=10bbea96c150de59e305b1aff956ca39" width="24" height="24" data-path="images/icons/providers/openai.svg">
    **Provider**: OpenAI

    DALL-E technology with high-quality generation and broad style support.
  </Card>

  <Card title="flux-pro-v1.1" icon="https://mintcdn.com/zerotwo/6HRSbZHwyN84Av-J/images/icons/providers/fal.svg?fit=max&auto=format&n=6HRSbZHwyN84Av-J&q=85&s=e9a14c0b9b8780f4104a365e9d190006" width="24" height="24" data-path="images/icons/providers/fal.svg">
    **Provider**: FAL

    Professional-quality images with advanced control and consistency.
  </Card>

  <Card title="imagen-4.0" icon="https://mintcdn.com/zerotwo/6HRSbZHwyN84Av-J/images/icons/providers/gemini.svg?fit=max&auto=format&n=6HRSbZHwyN84Av-J&q=85&s=f840a8ea8ca84cfbbeea67d191b535f5" width="24" height="24" data-path="images/icons/providers/gemini.svg">
    **Provider**: Google

    Latest Imagen technology with photorealistic quality and prompt following.
  </Card>

  <Card title="gemini-2.5-flash-image" icon="https://mintcdn.com/zerotwo/6HRSbZHwyN84Av-J/images/icons/providers/gemini.svg?fit=max&auto=format&n=6HRSbZHwyN84Av-J&q=85&s=f840a8ea8ca84cfbbeea67d191b535f5" width="24" height="24" data-path="images/icons/providers/gemini.svg">
    **Provider**: Google

    Fast multimodal image generation integrated with Gemini.
  </Card>

  <Card title="qwen-image" icon="https://mintcdn.com/zerotwo/6HRSbZHwyN84Av-J/images/icons/providers/qwen.svg?fit=max&auto=format&n=6HRSbZHwyN84Av-J&q=85&s=df430cb9716813c47ffbbb88203666a1" width="24" height="24" data-path="images/icons/providers/qwen.svg">
    **Provider**: Alibaba

    Multilingual prompts with excellent Asian aesthetics and styles.
  </Card>

  <Card title="grok-2-image" icon="https://mintcdn.com/zerotwo/6HRSbZHwyN84Av-J/images/icons/providers/grok.svg?fit=max&auto=format&n=6HRSbZHwyN84Av-J&q=85&s=d6ca5a0deee254a1c33182d1d6342430" width="24" height="24" data-path="images/icons/providers/grok.svg">
    **Provider**: X.AI

    Grok-powered generation with unique style and interpretation.
  </Card>
</CardGroup>

## Image Editing Models

Edit and manipulate existing images with AI.

* **gpt-image-1** (OpenAI): Advanced editing, inpainting, and manipulation
* **gemini-2.5-flash-image-preview** (Google): Multimodal editing with natural language
* **qwen-image-edit** (Alibaba): Qwen-powered editing with style control

## Choosing the Right Model

<AccordionGroup>
  <Accordion title="For Code Generation">
    **Best**: GPT-4o, GPT-4.1, Claude Sonnet 4

    These models excel at generating clean, documented code across multiple languages.
  </Accordion>

  <Accordion title="For Long Documents">
    **Best**: Claude Sonnet 4, Claude 3.5 Sonnet, Gemini 2.5 Pro

    Extended context windows make these ideal for analyzing lengthy documents.
  </Accordion>

  <Accordion title="For Real-Time Information">
    **Best**: Perplexity Sonar Pro, Grok-4, Gemini 2.5 Pro

    Built-in web access ensures current, factual information.
  </Accordion>

  <Accordion title="For Creative Writing">
    **Best**: GPT-4o, Claude Sonnet 4, GPT-4.1

    Strong language generation with creative flair and style control.
  </Accordion>

  <Accordion title="For Fast Responses">
    **Best**: Groq Llama 3.3, Gemini 2.5 Flash, GPT-4o-mini

    Ultra-low latency without sacrificing quality.
  </Accordion>

  <Accordion title="For Cost Efficiency">
    **Best**: GPT-4.1-nano, GPT-4o-mini, Mistral Small, Grok-3-mini

    Excellent performance at lower computational costs.
  </Accordion>

  <Accordion title="For Reasoning Tasks">
    **Best**: O3, O4-mini, DeepSeek Reasoner, Qwen Thinking

    Specialized for complex logic and mathematical problems.
  </Accordion>

  <Accordion title="For Multilingual Tasks">
    **Best**: Qwen series, Gemini 2.5 Pro, GPT-4o

    Exceptional performance across multiple languages.
  </Accordion>
</AccordionGroup>

<Tip>
  Start with GPT-4o or Claude Sonnet 4 for general tasks, then experiment with specialized models as you discover your specific needs.
</Tip>

## Model Comparison Table

| Provider             | Context Window | Speed      | Best Use Case           |
| -------------------- | -------------- | ---------- | ----------------------- |
| OpenAI GPT-5         | 128K           | Fast       | General, code, creative |
| Claude Sonnet 4      | 200K           | Medium     | Documents, analysis     |
| Gemini 2.5 Pro       | 1M+            | Fast       | Multimodal, video       |
| Mistral Large        | 128K           | Very Fast  | Privacy, efficiency     |
| DeepSeek             | 64K            | Medium     | Reasoning, research     |
| Groq Llama 3.3       | 32K            | Ultra Fast | Real-time apps          |
| Grok-4               | 128K           | Fast       | Current events          |
| Perplexity Sonar Pro | 127K           | Medium     | Research, citations     |

## Next Steps

<CardGroup cols={2}>
  <Card title="Switch Models" icon="shuffle" href="/getting-started/switch-models">
    Learn how to switch between models in conversations
  </Card>

  <Card title="Provider Details" icon="info" href="/providers/openai">
    Explore detailed documentation for each provider
  </Card>

  <Card title="Model Pricing" icon="dollar-sign" href="/account/subscriptions">
    Understand pricing and subscription tiers
  </Card>

  <Card title="Best Practices" icon="lightbulb" href="/prompts/overview">
    Learn effective prompting for different models
  </Card>
</CardGroup>