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

# AI Models

> Access to state-of-the-art AI models from leading providers

# AI Models

Pulze provides instant access to state-of-the-art AI models from all major providers. No need to set up individual accounts—we manage all AI integrations for you.

## Available Providers

<CardGroup cols={2}>
  <Card title="OpenAI" href="#openai">
    <div style={{textAlign: 'center', marginBottom: '10px'}}>
      <img src="https://mintcdn.com/pulzeai/iIc-koTvEb657C0a/_images/model-providers/openai.svg?fit=max&auto=format&n=iIc-koTvEb657C0a&q=85&s=b2388c9b8c93aa4350eaf8de4b0730f2" alt="OpenAI" width="60" height="60" data-path="_images/model-providers/openai.svg" />
    </div>

    GPT-5 family, o3 series, and more
  </Card>

  <Card title="Anthropic" href="#anthropic">
    <div style={{textAlign: 'center', marginBottom: '10px'}}>
      <img src="https://mintcdn.com/pulzeai/iIc-koTvEb657C0a/_images/model-providers/anthropic.svg?fit=max&auto=format&n=iIc-koTvEb657C0a&q=85&s=9543fa51b7797beed2199446bbb0dbc8" alt="Anthropic" width="60" height="60" data-path="_images/model-providers/anthropic.svg" />
    </div>

    Claude 4 and Claude 4.1 models
  </Card>

  <Card title="xAI" href="#xai">
    <div style={{textAlign: 'center', marginBottom: '10px'}}>
      <img src="https://mintcdn.com/pulzeai/iIc-koTvEb657C0a/_images/model-providers/xai.svg?fit=max&auto=format&n=iIc-koTvEb657C0a&q=85&s=d4b1fb8fd4342139663159e2e21535dc" alt="xAI" width="60" height="60" data-path="_images/model-providers/xai.svg" />
    </div>

    Grok 4 Fast - world's best model
  </Card>

  <Card title="Google" href="#google">
    <div style={{textAlign: 'center', marginBottom: '10px'}}>
      <img src="https://mintcdn.com/pulzeai/iIc-koTvEb657C0a/_images/model-providers/google.svg?fit=max&auto=format&n=iIc-koTvEb657C0a&q=85&s=6d1ea70db60dd220be6f943af4c5d141" alt="Google" width="60" height="60" data-path="_images/model-providers/google.svg" />
    </div>

    Gemini 2.5 Flash and Pro
  </Card>

  <Card title="AI21Labs" href="#ai21labs">
    <div style={{textAlign: 'center', marginBottom: '10px'}}>
      <img src="https://mintcdn.com/pulzeai/iIc-koTvEb657C0a/_images/model-providers/ai21labs.svg?fit=max&auto=format&n=iIc-koTvEb657C0a&q=85&s=9578ae62d561db723ee99399e4e9b190" alt="AI21Labs" width="60" height="60" data-path="_images/model-providers/ai21labs.svg" />
    </div>

    Jamba 1.7 architecture
  </Card>

  <Card title="Groq" href="#groq">
    <div style={{textAlign: 'center', marginBottom: '10px'}}>
      <img src="https://mintcdn.com/pulzeai/iIc-koTvEb657C0a/_images/model-providers/groq.svg?fit=max&auto=format&n=iIc-koTvEb657C0a&q=85&s=5ed8d41ad7a6ab8149b6737e267069c9" alt="Groq" width="60" height="60" data-path="_images/model-providers/groq.svg" />
    </div>

    Ultra-fast inference
  </Card>

  <Card title="Cohere" href="#cohere">
    <div style={{textAlign: 'center', marginBottom: '10px'}}>
      <img src="https://mintcdn.com/pulzeai/iIc-koTvEb657C0a/_images/model-providers/cohere.svg?fit=max&auto=format&n=iIc-koTvEb657C0a&q=85&s=47595ff9318fef0074820f8b275b6de8" alt="Cohere" width="60" height="60" data-path="_images/model-providers/cohere.svg" />
    </div>

    Enterprise language models
  </Card>

  <Card title="Fireworks" href="#fireworks">
    <div style={{textAlign: 'center', marginBottom: '10px'}}>
      <img src="https://mintcdn.com/pulzeai/iIc-koTvEb657C0a/_images/model-providers/fireworks.svg?fit=max&auto=format&n=iIc-koTvEb657C0a&q=85&s=9edbdf00f601ddef802df1cc98771baa" alt="Fireworks" width="60" height="60" data-path="_images/model-providers/fireworks.svg" />
    </div>

    Open-weight models
  </Card>
</CardGroup>

***

## OpenAI

Leading AI research company providing state-of-the-art models for various applications.

### GPT-5 Family

<AccordionGroup>
  <Accordion title="GPT-5" icon="star">
    **Flagship model** with massive 400K context window

    **Capabilities:**

    * Advanced reasoning and problem-solving
    * Complex code generation and debugging
    * Long-form content creation
    * Multi-step task execution
    * Agentic workflows

    **Specifications:**

    * **Context Window**: 400,000 tokens
    * **Multimodal**: Text only
    * **Best For**: Complex reasoning, long documents, agentic tasks
    * **Performance**: Highest intelligence across all tasks
  </Accordion>

  <Accordion title="GPT-5 Mini" icon="bolt">
    **Cost-efficient** model with strong reasoning

    **Capabilities:**

    * High-quality reasoning at lower cost
    * Code generation and analysis
    * Content creation
    * Data analysis

    **Specifications:**

    * **Context Window**: 128,000 tokens
    * **Multimodal**: Text only
    * **Best For**: Cost-sensitive applications with strong reasoning needs
    * **Performance**: Fast, affordable, maintains strong reasoning
  </Accordion>

  <Accordion title="GPT-5 Nano" icon="gauge-high">
    **Fastest** and most affordable option

    **Capabilities:**

    * Quick responses
    * Basic reasoning
    * Simple code tasks
    * General conversation

    **Specifications:**

    * **Context Window**: 32,000 tokens
    * **Multimodal**: Text only
    * **Best For**: High-throughput applications, real-time responses
    * **Performance**: Ultra-fast, budget-friendly, good for production scale
  </Accordion>

  <Accordion title="GPT-5 Codex" icon="code">
    **Optimized** for agentic coding environments

    **Capabilities:**

    * Advanced code generation
    * Multi-file code editing
    * Debugging and optimization
    * Test generation
    * Code review and analysis

    **Specifications:**

    * **Context Window**: 200,000 tokens
    * **Multimodal**: Text and code
    * **Best For**: Code generation, debugging, development workflows
    * **Performance**: Enhanced code understanding, multi-language support
  </Accordion>

  <Accordion title="GPT-5 Chat" icon="comments">
    **ChatGPT-optimized** variant

    **Capabilities:**

    * Natural conversation
    * Context retention
    * Personality consistency
    * Multi-turn dialogue

    **Specifications:**

    * **Context Window**: 128,000 tokens
    * **Multimodal**: Text only
    * **Best For**: Conversational applications, chatbots
    * **Performance**: Tuned for dialogue, natural conversations
  </Accordion>
</AccordionGroup>

### Research Models

<AccordionGroup>
  <Accordion title="o3-deep-research" icon="magnifying-glass-chart">
    **Research model** for complex multi-step research tasks

    * **Best For**: Deep analysis, multi-step reasoning, comprehensive research
    * **Key Features**: Extended thinking time, thorough analysis, citation-rich outputs
  </Accordion>
</AccordionGroup>

***

## Anthropic

AI safety company known for helpful, harmless, and honest AI systems.

<AccordionGroup>
  <Accordion title="Claude Opus 4.1" icon="crown">
    **Very capable** model with highest intelligence

    * **Best For**: Most demanding tasks requiring maximum capability
    * **Key Features**: Superior reasoning, excellent at complex analysis
  </Accordion>

  <Accordion title="Claude Sonnet 4.5" icon="code">
    **Best** for complex agents and coding tasks

    * **Best For**: Agentic workflows, software development, automation
    * **Key Features**: Excellent tool use, reliable coding, strong reasoning
  </Accordion>

  <Accordion title="Claude Sonnet 4.0" icon="file-lines">
    **Long-context** model with 1M token window

    * **Context**: 1,000,000 tokens (upgraded from 200K)
    * **Best For**: Extremely long documents, entire codebases
    * **Key Features**: 5x increase in context length, maintains quality at scale
  </Accordion>
</AccordionGroup>

***

## xAI

Advanced AI models with massive context windows and state-of-the-art performance.

<AccordionGroup>
  <Accordion title="Grok 4 Fast" icon="trophy">
    **World's best model** - currently leading all benchmarks

    * **Context**: 2,000,000 tokens
    * **Best For**: Most demanding AI tasks
    * **Key Features**:
      * Leading all benchmarks as best model in the world
      * Massive 2M context window
      * Advanced reasoning capabilities
      * Complex problem solving
  </Accordion>

  <Accordion title="Grok Code Fast" icon="code">
    **Optimized** for agentic coding tasks

    * **Best For**: Advanced development workflows, code generation
    * **Key Features**: State-of-the-art coding, rapid development
  </Accordion>
</AccordionGroup>

***

## Google

Advanced multimodal models with strong reasoning capabilities.

<AccordionGroup>
  <Accordion title="Gemini 2.5 Flash" icon="bolt">
    **Best price/performance** with thinking capabilities

    * **Best For**: Production applications, cost-sensitive deployments
    * **Key Features**:
      * Built-in reasoning ("thinking")
      * Fast inference
      * Excellent value
      * Multimodal (text, images, video)
  </Accordion>

  <Accordion title="Gemini 2.5 Pro" icon="brain">
    **Advanced reasoning** with 1M+ context window

    * **Context**: 1,000,000+ tokens
    * **Best For**: Complex reasoning, long documents, research
    * **Key Features**: Superior reasoning, extended context, multimodal
  </Accordion>
</AccordionGroup>

***

## AI21Labs

Novel Mamba-Transformer hybrid architecture for enterprise applications.

<AccordionGroup>
  <Accordion title="Jamba Large 1.7" icon="building">
    **Enterprise-scale** performance with 256K context

    * **Context**: 256,000 tokens
    * **Architecture**: Mamba-Transformer hybrid
    * **Best For**: Enterprise applications, long documents
    * **Key Features**: Unique architecture, efficient processing
  </Accordion>

  <Accordion title="Jamba Mini 1.7" icon="gauge">
    **Efficient** with 256K context window

    * **Context**: 256,000 tokens
    * **Best For**: Cost-effective enterprise applications
    * **Key Features**: Smaller footprint, maintains capabilities
  </Accordion>
</AccordionGroup>

***

## Groq

Ultra-fast inference with integrated tool orchestration.

<AccordionGroup>
  <Accordion title="Compound Systems" icon="gears">
    **Integrated tool orchestration**

    * **Best For**: Multi-tool workflows, complex automation
    * **Key Features**:
      * Built-in web search
      * Code execution
      * Browser automation
      * Coordinated tool use
  </Accordion>

  <Accordion title="OpenAI GPT-OSS Models" icon="code-branch">
    **Open-weight** models (20B & 120B)

    * **Variants**: 20B and 120B parameters
    * **Best For**: Transparency, customization, research
    * **Key Features**: Open-source ecosystem, full access
  </Accordion>

  <Accordion title="Llama 4 Models" icon="meta">
    **Meta's latest** multimodal models

    * **Llama 4 Maverick**: Flagship multimodal model
    * **Llama 4 Scout**: Balanced performance and efficiency
    * **Best For**: Multimodal applications, versatile tasks
  </Accordion>
</AccordionGroup>

***

## Cohere

Enterprise-focused language models for production applications.

<AccordionGroup>
  <Accordion title="Command A 03-2025" icon="terminal">
    **Enterprise language model**

    * **Context**: 256,000 tokens
    * **Parameters**: 111 billion
    * **Best For**: Enterprise applications, command and control
    * **Key Features**: Production-ready, enterprise support
  </Accordion>
</AccordionGroup>

***

## Fireworks

Fast inference platform for open-weight models.

<AccordionGroup>
  <Accordion title="GPT-OSS Series" icon="fire">
    **Multiple parameter sizes**

    * **Best For**: Open-source deployments, customization
    * **Key Features**: Fast inference, open weights, flexible
  </Accordion>

  <Accordion title="DeepSeek V3.1" icon="magnifying-glass">
    **Latest reasoning models**

    * **Best For**: Advanced reasoning tasks
    * **Key Features**: Strong reasoning, open architecture
  </Accordion>
</AccordionGroup>

***

## Model Selection Guide

### By Use Case

**Reasoning & Analysis**

* xAI Grok 4 Fast (best overall)
* OpenAI GPT-5
* Anthropic Claude Opus 4.1
* Google Gemini 2.5 Pro

**Coding & Development**

* Anthropic Claude Sonnet 4.5
* OpenAI GPT-5 Codex
* xAI Grok Code Fast

**Cost-Performance**

* Google Gemini 2.5 Flash
* OpenAI GPT-5 Nano
* OpenAI GPT-5 Mini

**Long Context**

* Anthropic Claude Sonnet 4.0 (1M)
* Google Gemini 2.5 Pro (1M+)
* xAI Grok 4 Fast (2M)
* AI21Labs Jamba (256K)

**Enterprise Production**

* Cohere Command A
* AI21Labs Jamba Large
* Google Gemini 2.5 Flash

### Custom Routing

Don't want to choose manually? Use [custom routers](/features/custom-routers) to automatically select the best model based on your requirements.

***

## Model Lifecycle

<Note>
  Models are regularly updated with new versions and capabilities. Some older models may be deprecated over time.
</Note>

<CardGroup cols={2}>
  <Card title="Deprecations" icon="circle-exclamation" href="/models/deprecations">
    View deprecated models and migration guidance
  </Card>

  <Card title="Changelog" icon="clock-rotate-left" href="/changelog">
    See latest model updates and releases
  </Card>
</CardGroup>

***

## Getting Started

<Steps>
  <Step title="Choose Your Use Case">
    Identify what you need the model for (reasoning, coding, production, etc.)
  </Step>

  <Step title="Select Model">
    Pick from our recommendations or use custom routing
  </Step>

  <Step title="Test Performance">
    Run evaluations against your datasets
  </Step>

  <Step title="Deploy">
    Use in your spaces or via API
  </Step>

  <Step title="Monitor">
    Track performance and adjust as needed
  </Step>
</Steps>

## Next Steps

<CardGroup cols={2}>
  <Card title="Custom Routers" icon="route" href="/features/custom-routers">
    Automatically select the right model
  </Card>

  <Card title="Evaluations" icon="clipboard-check" href="/features/evaluations">
    Test models against your standards
  </Card>

  <Card title="API Reference" icon="code" href="/api-reference/chat/chat-completions">
    Integrate models into your applications
  </Card>

  <Card title="Model Deprecations" icon="circle-exclamation" href="/models/deprecations">
    Stay updated on model lifecycle
  </Card>
</CardGroup>
