What happens when you send a message to an LLM (the API call explained)
This article explains that Large Language Models are stateless and do not inherently remember previous interactions. To simulate a continuous conversation, the entire message history must be sent with every new API request. The text details how the messages array structures this data using specific roles to define the context for the model.
- ▪LLMs are stateless prediction machines that start fresh with no memory of previous messages on every API call.
- ▪The illusion of conversation memory is created by sending the full transcript of the interaction history with each new request.
- ▪The messages array is a structured list where each entry is labeled with a role to indicate who sent the content.
- ▪The three primary roles in the messages array are user, assistant, and system, which define the participants and rules of the interaction.
Hacker News (AI / LLM) files mainly under ai. We currently carry 5,386 of its stories.
Story provenance
Source · retrieval · rights · ranking — open for full record
inspect →
Story provenance
Attribution is not the same as permission. This drawer separates discovery metadata, excerpts, WeSearch-generated summaries, reuse status, and whether the publisher receives the visit. Nothing here claims a legal grant the publisher has not made.
Record
| Original publisher | Hacker News (AI / LLM) |
| Canonical URL | https://heymeraki.substack.com/p/aie_20-inside-the-llm-api-call |
| Publication time | Thu, 17 Sep 2026 19:09:33 +0000 |
| Retrieval time | 2026-09-17T19:28:44.416Z |
| Last seen | 2026-09-17T19:28:44.416Z |
| Headline source | Publisher (no WeSearch rewrite) |
| Excerpt source | publisher body |
| Excerpt method | First ~120 words (~800 chars) of extracted publisher body, fair-use limited. |
| Summary | WeSearch · cerebras-chat (WeSearch summarizer) |
| Summary source text | contentText |
| Citation coverage | Summary is a WeSearch-generated derivative; primary citation is the original publisher URL. |
| Cluster | 2pEGTgPS-ubQ · 1 stories |
| Cluster logic | Grouped by semantic title/content similarity across sources within a rolling window. Same-publisher template collisions are excluded from coverage comparison. |
| Ranking reason | Story pages are not engagement-ranked. Hub feeds use recency, with optional source-diversified chronological ordering (cap consecutive stories per source). No personalized ranking. |
| Publisher visit | Yes — open original |
| Substitutes article? | No — link-out required for full text |
Rights status (four layers)
WeSearch handling by dimension
| Indexing | May the item be indexed (stored, ranked, made findable)? | Allowed |
| Snippet | May a short excerpt of the publisher's text be shown? | Allowed |
| AI summary | May WeSearch generate its own short summary of the article? | Limited |
| Retrieval / RAG | May the content be exposed for third-party retrieval-augmented generation? | Not asserted |
| Model training | May the content be used to train AI models? | Not asserted |
| Commercial reuse | May the content be reused commercially? | Not permitted |
Basis: Derived from the published RSS/Atom feed. Contact: [email protected]. Reviewed: 2026-07-24.
Opening excerpt (first ~120 words) tap to expand
aie_2.0: inside the LLM API callwhat happens when you hit send to an LLMmerakiSep 17, 20261ShareIn the last lesson, we established that an LLM (Large Language Model) is a prediction machine. It does not know things, it simply predicts the most likely next word based on all that it was trained on. We also established that it is stateless, no memory, every time you send a message, it starts fresh with no memory of the previous messages.That raised a question for us. If it has no memory, how is it that a conversation with Claude or ChatGPT feels like a real conversation with callouts and references to previous messages? How is it that it seems to remember what you said five messages ago?The answer is what this lesson is about. It has to do with how the API call is structured.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Hacker News (AI / LLM).
Discussion
0 commentsMore from Hacker News (AI / LLM)
