AI LLM APIs (Foundation Models)
Access the world's most capable language models via API to power your product's AI features — from chatbots and content generation to complex reasoning and data extraction. These platforms handle the model infrastructure so you focus on building, not running GPU servers.
ChatGPT
Engage in dynamic conversations, debug code, and generate creative content with advanced AI.
Hugging Face
The AI community platform for hosting, sharing, and running open machine learning models
Amazon Bedrock
The fully managed AWS platform for building generative AI applications and agents at production scale
Picsart
The AI creative platform for 130M+ creators. Turn any idea into scroll-stopping content.
HeyGen
Transform ideas into stunning AI videos with realistic avatars and multi-language support.
Inworld AI
Realtime TTS, STT, and LLM routing infrastructure for consumer-scale voice AI
Google Gemini API
Build with Google's multimodal Gemini models via API and AI Studio
Groq
The fastest inference cloud for open-source LLMs, powered by custom LPU chips
LangChain
The open-source framework and platform for building reliable AI agents
NVIDIA ACE
Bring digital humans to life with generative AI microservices for speech, intelligence, and animation
Daloopa
Audit-ready financial data infrastructure powering AI-driven investment research
Stable Diffusion
The open-source AI image generation model — run locally for free or access via API, with maximum creative control
Fireworks AI
High-performance training and inference platform for open-source AI models
LLM APIs & Foundation Model Providers
Foundation model APIs let developers send text (and now images, audio, and documents) to a large language model and get intelligent responses back — the same underlying capability powering tools like ChatGPT and Claude, but accessible via API so you can build it into your own product.
The main providers worth knowing
- OpenAI — GPT models plus image generation, audio transcription, and embeddings.
- Anthropic — Claude models, noted for long context windows and instruction following.
- Google — Gemini models with strong multimodal capabilities.
- Mistral — efficient open-weight models available via API or self-hosted.
What to compare between providers
Context window size, pricing per token, latency, and rate limits all affect which provider fits your use case. For production applications, reliability and uptime history matter too — it's worth checking how each provider has handled outages and how they communicate about them.
Compare AI LLM APIs (Foundation Models) Tools
Direct side-by-side feature and pricing evaluations
Also explore in AI Developer APIs & Platforms
AI Agent & Orchestration Frameworks
Build AI applications that do more than chat — agents that search the web, run code, query databases, call APIs, and hand off tasks between specialized sub-agents. These frameworks give you the building blocks for multi-step AI workflows without building the orchestration layer from scratch.
AI Cloud ML Platforms
Build, train, deploy, and monitor machine learning models on enterprise-grade cloud infrastructure from AWS, Google, Microsoft, and IBM. These platforms handle the heavy lifting of data management, model training at scale, and deployment pipelines — so your ML team focuses on the models, not the infrastructure.
AI Computer Vision & Speech APIs
Add the ability to see, read, and listen to your applications — via APIs for image recognition, OCR, object detection, speech-to-text, and speaker identification. These are the building blocks behind AI apps that process documents, analyze photos, or transcribe audio at scale.
AI Model Hosting & Open-Source Model APIs
Run open-source models like Llama, Mistral, and Qwen at scale without managing your own GPU infrastructure — through APIs that feel familiar but give you access to open-weight models you can customize, fine-tune, or deploy under your own terms.
AI Vector Databases & RAG Infrastructure
Power semantic search and retrieval-augmented generation (RAG) apps with a database built for AI embeddings. Store and query millions of vectors fast — the infrastructure layer behind AI applications that need to search documents, memories, or knowledge bases by meaning, not just keywords.