AI Glossary
What is Inference?
Training builds the model; inference uses it. Every chat reply, code completion, or generated image is inference. Its cost scales with model size and tokens processed, which drives the industry pattern of cheap small models for high-volume tasks and expensive frontier models for hard ones.
News momentum
Mentions in recent AI news titles and summaries, refreshed from the ingestion stream.
231
-93 vs prior 7d
- Someone hacked Prime Intellect's inference servers and replaced DeepSeek V4.1 Flash with an abliterated version stripped of safety refusals…
bluesky-ai · 2h
- What Would the Ideal LLM Inference Chip Look Like? [video]
hn-ai · 14h
- From Closed to Open-Weight LLMs: Choosing the Inference Stack for Voice AI
hn-ai · 18h
- 180B Parameters, 32GB of RAM: Pushing Moe Inference to the Limit
hn-ai · 1d