Volume III · Field Identification
The identification field guide
Every frontier family leaves linguistic and behavioural fingerprints. The eleven topics below are the field methods for distinguishing one from another by output alone.
Identifying a model from its prose alone
Every frontier family leaves linguistic fingerprints. Here is the field guide.
Context-window claims, real and theoretical
Published context lengths and effective context lengths are different numbers. Here is how to test the gap.
Pricing models and how to forecast a bill
Per-token pricing is the simple part. Reasoning tokens, cached prefixes, and tool-call billing are where surprises live.
Tool-use schemas and the JSON-adherence problem
Every frontier model claims tool-use support. Adherence-rate on schemas tells a different story.
Agent loops: what makes one converge and another stall
An agent loop is a planner, a tool layer, and a state buffer. Each of the three has classic failure modes.
Prompt caching, the 90%-discount most operators don't use
Anthropic, OpenAI, and Gemini all offer cached-prefix discounts. The architectures that take advantage of them look different from the ones that don't.
Extended thinking: when to spend reasoning tokens and when not
Reasoning effort is a knob, not a switch. The cost-to-correctness curve is real and operators should sketch it before turning it up.
Refusals: what models won't do and why
Each frontier family has a different refusal training. A workload that one model refuses, another will attempt; a workload one attempts, another will quietly hallucinate.
Fine-tuning: what's possible, what's worth it
Fine-tuning closed models is now table-stakes. Whether it's worth your money is a different question.
Deployment: APIs, dedicated, self-hosted
Where the model runs changes everything about cost, latency, privacy, and operational burden.
Benchmarks: which ones to trust, which to discount
Every benchmark eventually gets gamed. The trick is knowing which are useful in their current form.