Build assistants that reason, write code, and use tools with the Claude API.
AI modelsLanguage modelsTool use
Documentation Source reviewed
AI model APIs
These are providers and inference APIs, not individual model versions. Compare supported models, input and output types, data policies and deployment options in each provider’s documentation.
Editorial selection · Summaries reviewed · Our methodologyChoose your building blocks
Build enterprise AI with language models, embeddings, and reranking for retrieval workflows.
AI modelsEmbeddingsReranking
Documentation Source reviewed
Run supported language and audio models through Groq’s inference API.
AI modelsInferenceAudio
Documentation Source reviewed
Build applications with language, image and audio models through the OpenAI API.
AI modelsLanguage modelsMultimodal
Documentation Source reviewed
Generate content and analyze multimodal inputs with the Gemini API.
AI modelsLanguage modelsMultimodal
Documentation Source reviewed
Access Mistral models through an API for text generation and AI application development.
AI modelsLanguage modelsInference
Documentation Source reviewed
Integrate DeepSeek models through its chat completion API.
AI modelsLanguage modelsChat completions
Documentation Source reviewed
Build applications with Grok models through the xAI API.
AI modelsLanguage modelsInference
Documentation Source reviewed
Run supported open models through hosted inference APIs and dedicated endpoints.
AI modelsOpen modelsInference
Documentation Source reviewed
Serve supported open models using hosted inference and deployment APIs.
AI modelsOpen modelsInference
Documentation Source reviewed
Access supported language models through the Cerebras inference API.
AI modelsLanguage modelsInference
Documentation Source reviewed
Run machine-learning models through an API, including models for images, video and language.
AI modelsMultimodalInference
Documentation Source reviewed
Generate images and videos through an asynchronous API, with polling and webhook-based completion workflows.
AI modelsImage generationVideo generation
Documentation Source reviewed
Build model training and sampling workflows with Thinking Machines’ Tinker API and Python SDK.
AI modelsFine-tuningLanguage models
Documentation Source reviewed
What to evaluate
- Evaluate models on representative tasks from your application.
- Check the current model catalog and API capabilities before integrating.
- Review data retention, regional availability and usage costs with each provider.
Provider documentation
Read the original sources for current capabilities, setup requirements, and usage terms. These selections are starting points, not a performance ranking.
| Provider | Use case | Primary source |
|---|---|---|
| Anthropic | Language models, Tool use | Anthropic documentation ↗ |
| Cohere | Embeddings, Reranking | Cohere documentation ↗ |
| Groq | Inference, Audio | Groq documentation ↗ |
| OpenAI | Language models, Multimodal | OpenAI documentation ↗ |
| Google Gemini | Language models, Multimodal | Google Gemini documentation ↗ |
| Mistral AI | Language models, Inference | Mistral AI documentation ↗ |
| DeepSeek | Language models, Chat completions | DeepSeek documentation ↗ |
| Grok API (xAI) | Language models, Inference | Grok API (xAI) documentation ↗ |
| Together AI | Open models, Inference | Together AI documentation ↗ |
| Fireworks AI | Open models, Inference | Fireworks AI documentation ↗ |
| Cerebras | Language models, Inference | Cerebras documentation ↗ |
| Replicate | Multimodal, Inference | Replicate documentation ↗ |
| Higgsfield | Image generation, Video generation | Higgsfield documentation ↗ |
| Tinker (Thinking Machines) | Fine-tuning, Language models | Tinker (Thinking Machines) documentation ↗ |