Cheap models for labelling and filtering
Small and decision models for high-volume, simple tasks.
- Wednesday, November 11, 2026 · F3 Cheap small models for high-volume tasks: Meta Muse Ember
Meta's smallest model runs at $0.10 / $0.50 per 1M tokens, aimed at high-volume agent tasks (tool calls, routing, data extraction), putting it on par with Haiku 5.5. The question: is it worth…
- Thursday, November 12, 2026 · K4 A decision model instead of an LLM: Amazon labels 44 times cheaper
In an internal ticketing system, Amazon replaced Claude Sonnet 4.6 with its own Strands Decider model, which doesn't generate text but gives probabilities for predefined labels. Accuracy stayed the…
- Friday, November 13, 2026 · K2 Gemini Decisions API: cheap labeling for high-volume text
Google's new model (Gemini Decision 1) doesn't write text; it gives probabilities for predefined questions: yes/no, multiple choice (2–255 options) or rubric scoring. In one request it evaluates up…
- Saturday, November 14, 2026 · F3 Perplexity Decider: another classifier model
Decider v1.3 and Decider Flash v1.3 return probabilities for yes/no and multi-class labelling, for text and images. Alongside the Gemini Decisions API (Nov 13 / K2), this is a second cloud option for…