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

# Build with Fireworks AI

> Fast inference and training for open-source models on Fireworks AI, covering serverless APIs, dedicated deployments, fine-tuning, and multi-modal workflows.

Fireworks AI is the fastest platform for building with open source AI models. Get production-ready inference and training with best-in-class speed, cost and quality.

## Get started in minutes

> [card] [Start fast with Serverless](https://fireworks.ai/docs/getting-started/quickstart.md)
>
> Use popular models instantly with pay-per-token pricing. Perfect for quality vibe testing and prototyping.

> [card] [Deploy models & autoscale on dedicated GPUs](https://fireworks.ai/docs/getting-started/ondemand-quickstart.md)
>
> Deploy with high performance on dedicated GPUs with fast autoscaling and minimal cold starts. Optimize deployments for speed and throughput.

> [card] [Train models for best quality](https://fireworks.ai/docs/fine-tuning/finetuning-intro.md)
>
> Boost model quality with supervised and reinforcement fine-tuning of models up to 1T+ parameters. Start training in minutes, deploy immediately.

> [card] [Connect your coding tools with Nexus](https://fireworks.ai/docs/nexus.md)
>
> Connect coding tools with Nexus CLI, choose Fireworks models directly or through Nexus Auto, and manage supported serverless usage.

> [!TIP]
>
> Not sure where to start? First, pick the right model for your use case with our [**model selection guide**](https://fireworks.ai/docs/guides/recommended-models.md). Then choose [**Serverless**](https://fireworks.ai/docs/getting-started/quickstart.md) to prototype quickly, move to [**Deployments**](https://fireworks.ai/docs/getting-started/ondemand-quickstart.md) to optimize and run production workloads, or use [**Training**](https://fireworks.ai/docs/fine-tuning/finetuning-intro.md) to improve quality.
>
> New to AI or Fireworks? Look up any term in the [**Glossary**](https://fireworks.ai/docs/getting-started/glossary.md).
>
> Need help optimizing deployments, training models, or setting up production infrastructure? [Talk to our team](https://fireworks.ai/company/contact-us) - we'll help you get the best performance and reliability.

## What you can build

> [card] [100+ Supported Models](https://fireworks.ai/models)
>
> Text, vision, audio, image, and embeddings

> [card] [Migrate from OpenAI](https://fireworks.ai/docs/tools-sdks/openai-compatibility.md)
>
> Drop-in replacement for inference and training — same API, same SFT data format

> [card] [Function Calling](https://fireworks.ai/docs/guides/function-calling.md)
>
> Connect models to tools and APIs

> [card] [Structured Outputs](https://fireworks.ai/docs/structured-responses/structured-response-formatting.md)
>
> Reliable JSON responses for agentic workflows

> [card] [Vision Models](https://fireworks.ai/docs/guides/querying-vision-language-models.md)
>
> Analyze images and documents

> [card] [Embeddings & Reranking](https://fireworks.ai/docs/guides/querying-embeddings-models.md)
>
> Use embeddings & reranking in search & context retrieval

> [card] [Batch Inference](https://fireworks.ai/docs/guides/batch-inference.md)
>
> Run async inference jobs at scale, faster and cheaper

## Resources & help

> [card] [Which model should I use?](https://fireworks.ai/docs/guides/recommended-models.md)
>
> Find the best model for your use case

> [card] [Cookbook](https://github.com/fw-ai-external/cookbook)
>
> Code examples and tutorials

> [card] [API Reference](https://fireworks.ai/docs/api-reference/introduction.md)
>
> Complete API documentation

> [card] [Discord Community](https://discord.gg/fireworks-ai)
>
> Ask questions and get help from developers

> [card] [Security & Compliance](https://trust.fireworks.ai/)
>
> SOC 2, HIPAA, and audit reports

> [card] [System Status](https://status.fireworks.ai/)
>
> Check service uptime

> [card] [Talk to Sales](https://fireworks.ai/company/contact-us)
>
> Talk to our team


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.