What is an LLM API, and which major provider APIs should an engineer know in 2026?

beginner

Re-explain

An LLM API is a hosted large language model you call over HTTPS. You send a prompt plus settings and get generated text back. The model is stateless, so you resend the conversation or let the API store it.

The first-party APIs to know:

Anthropic's Messages API, POST /v1/messages.
OpenAI's Responses API, POST /v1/responses. Chat Completions still works.
Google's Gemini API, where the Interactions API is now recommended over generateContent.

Cloud platforms resell the same models: Amazon Bedrock, Microsoft Foundry, and Google's Gemini Enterprise Agent Platform, formerly Vertex AI. They add enterprise billing, regions, and cloud identity and access management (IAM).

Many smaller providers ship an OpenAI-compatible endpoint, so one client library often reaches Mistral, DeepSeek or self-hosted models.

Rewriting in plainer words…

This answer doesn't lend itself to a diagram - it reads best . No credits were charged.

Why there's no diagram: “”

The interactive diagram is below the answer - jump to diagram ↓ · Below it, the related concept . Jump to it ↓

The diagram below the answer is the concept . Jump to it ↓

Tailored explanation · switch back to · ·

What should the new diagram focus on?

Guided practice for LLM APIs & Integration One question at a time. Answer out loud, get graded, see what you missed.

How well did you know this?

AI:

Saved in this browser - sign in to keep your review list.

How should your speech become text?

Listening… your words appear above as you speak - tap Stop when you're done.

Recording · cr - tap Stop & transcribe when you're done.

Transcribing with AI…

Voice:

Keep going - a few more words and AI can grade it.

Interview lens How interviewers actually use this question

Likely follow-ups

If the model is stateless, how does a multi-turn chat keep context, and what does that cost you? Resending the full message history every turn, input tokens growing each turn, and the option of letting the API store the conversation server-side.
How does OpenAI's Responses API differ from Chat Completions? Server-side conversation state, built-in tools, and why new OpenAI work defaults to Responses while Chat Completions stays supported.
Beyond field names, what actually differs between the Anthropic, OpenAI and Gemini request shapes? Where the system prompt lives, how roles and content blocks are structured, and how tool calls and their results come back.
When would you call Claude through Bedrock instead of Anthropic's API directly? Existing AWS billing and IAM, region or data residency needs, weighed against possible feature lag and separate quotas on the cloud platform.
Many providers are OpenAI-compatible. Why not build everything on that one format? Compatibility covers the common shape. Caching, tool use and reasoning controls differ per provider and get lost behind a lowest-common-denominator layer.

What you can say

An LLM API is a hosted large language model that you call over HTTPS with a prompt and some settings, and it returns generated text.
The model is stateless, so for a chat you either resend the whole conversation every turn or let the API store it.
The first-party APIs I'd know are Anthropic's Messages API, OpenAI's Responses API, and Google's Gemini API, where the Interactions API is now recommended over generateContent.
Cloud platforms resell the same models: Amazon Bedrock, Microsoft Foundry, and Google's Gemini Enterprise Agent Platform, formerly Vertex AI.
Teams pick those for enterprise billing, region choice, and access control through the IAM they already run in that cloud.
Many smaller providers ship an OpenAI-compatible endpoint, so one client often reaches Mistral, DeepSeek or a self-hosted model, though not every parameter behaves the same.

Weak answers to avoid

Lists providers but never defines the API The question has two halves. Say first that it's a stateless HTTPS call with a prompt and settings, then name the APIs.
Knows only Chat Completions and generateContent That reads as knowledge from a few years ago. Name OpenAI's Responses API and Gemini's Interactions API, and note that Chat Completions still works.
Thinks the model remembers the conversation The model is stateless. You resend history or opt into the API storing it, and resent history grows your input tokens every turn.
Treats Bedrock or Foundry as different models They resell the same models. The reasons to use them are enterprise billing, regions and cloud IAM, not a different model.

Interview lens

Likely follow-ups, what you can say, and the weak answers to avoid.

Want a quick review of the fundamentals? See the LLM APIs & Integration cheatsheet.

← Back to all LLM APIs & Integration questions

Pro · $10/mo

45 of 52 LLM APIs & Integration answers are in Pro.

Full answers, code samples, and AI explanations that go simpler or deeper. Cancel anytime.

Full answers + code
AI explanations, simpler or deeper
1,000 AI credits / month
Cancel anytime

Get Pro → See plans & pricing →