Home / LLM APIs

Best LLM APIs

The best LLM API for coding and agentic apps is the Anthropic API (Claude). The OpenAI API has the broadest ecosystem and modalities, and the Google Gemini API is best for long context and has a free tier for experimenting. If you want to switch between hundreds of models with one key, use OpenRouter; if latency matters most, use Groq.

Model APIs to power your app. Prices last verified Oct 22, 2026.

LLM APIs rankings

1. #1

   Anthropic API

   paid

   Claude models via the Claude Platform.

   Best for: Coding and agentic workloads

   Pricing: Pay-as-you-go per token (e.g. Sonnet 5.5 $2/$10 per MTok in/out)
   Visit ↗
2. #2

   OpenAI API

   paid

   GPT models, images, audio and more.

   Best for: Broadest ecosystem and modalities

   Pricing: Pay-as-you-go per token
   Visit ↗
3. #3

   Google Gemini API

   freemium

   Gemini models via AI Studio and the Gemini API.

   Best for: Long context and free tier experimentation

   Pricing: Free tier with rate limits in AI Studio; pay-as-you-go for higher limits
   Visit ↗
4. #4

   OpenRouter

   freemium

   One API for hundreds of models.

   Best for: Switching models without rewiring

   Pricing: Free models (rate-limited) + pay-as-you-go credits
   Visit ↗
5. #5

   xAI API

   paid

   Grok models via API.

   Best for: Grok models in your app

   Pricing: Usage-based; from $2 per 1M input tokens
   Visit ↗
6. #6

   Groq

   freemium

   Very fast inference for open models.

   Best for: Latency-sensitive apps

   Pricing: Free tier (rate-limited); pay-as-you-go Developer tier
   Visit ↗
7. #7

   Mistral AI

   freemium

   Open and commercial models from Mistral.

   Best for: European hosting and open weights

   Pricing: Free plan (Le Chat/Vibe); Pro and pay-as-you-go API
   Visit ↗
8. #8

   Together AI

   paid

   Inference and fine-tuning for open models.

   Best for: Running open-weight models at scale

   Pricing: Pay-as-you-go per token
   Visit ↗