AI Model Reports
AI Model Research Library
Market research on model performance, adoption, economics, and production choices.
Selected reports
8
articles
Families
10
Major model families
Industry sources
11
Source Landscape
Signal types
4
coverage areas
Articles
Selected model research
Leaderboard signals, cost per task, fallback, agents, image, video, and regional model families.
AI Infrastructure Arms Race
A wide-angle look at compute, energy, open models, model data, routing, and agent infrastructure.
Open article
How Developers Should Read LLM API Leaderboards
A practical guide to reading usage, quality, speed, price, context, and task-fit rankings side by side.
Open article
AI Model Leaderboard Watch
A dated market read on model ranking movement, pricing context, and current model-selection signals.
Open article
Cost Per Task Beats Cost Per Token
Why output length, retries, cache behavior, and failure modes can change the real price of a model choice.
Open article
When Semantic Cache Returns the Wrong Answer
A production-oriented cautionary story about embedding similarity, template-heavy prompts, and cache correctness.
Open article
Agent Model Fallback Routing
How to think about reliability, spend, primary models, and backup models for agent workloads.
Open article
Building AI Agents With Multiple Models
A practical architecture guide for splitting agent work across models with different strengths.
Open article
Async Image Generation Tasks
Long-running image generation with task creation, polling, completion, and final assets.
Open article
