ModelsThe Qwen Family Explained: The Models You Can Own
Qwen is the tier that runs on hardware you have. What the 3.6 generation offers, why dense matters, and how to read a family with many sizes.
ReadNotes on models, agents, and what inference costs
Practical writing for developers building with large language models — how they work, how to pick one, and how to keep the bill predictable.
Models
Cost & PricingThe Real Cost of Migrating Between Model Providers
Changing the base URL takes an afternoon. Re-tuning prompts, rerunning evals and dual-running takes weeks. A worked estimate structure you fill in with your own numbers.
Read
Cost & PricingThe Real Cost of Open vs Closed Models
Open weights cost nothing to download and plenty to run. Here is how licence, hosting and switching costs actually compare, with current per-token rates.
Read
Cost & PricingThe ROI of Flat-Rate Billing for AI Development Work
The break-even maths for flat-rate AI plans, why usage variance decides the answer more than the average does, and the workloads where metered billing still wins.
Read
Cost & PricingThe Self-Hosting Break-Even Point, With Honest Inputs
Self-hosting beats per-token pricing above some volume. Working out where that point sits once engineering time and idle capacity are included.
ReadCost & PricingThe Token Cost of Long Context: When 1M Is Worth Filling
A million-token window makes enormous prompts possible, not advisable. What filling it costs across model tiers, and when retrieval wins instead.
Read
Cost & PricingThe Token Cost of Reasoning Models: Paying for Hidden Output
Reasoning models emit tokens you never see and are billed for at output rates. How much that adds, and when the accuracy is worth it.
ReadCost & PricingThe Token Cost of Tool Calls: Definitions Bill Every Turn
Tool definitions and results are prompt content, charged on every turn of every session. What a tool actually costs over a run, and how to shrink it.
Read
ModelsThe Z.ai GLM Family Explained: What GLM-5.2 Changes
GLM-5.2 pairs MIT licensing with a 128K output ceiling and two reasoning modes. What the family offers, and the workloads it is uniquely good at.
Read
FundamentalsThroughput vs Latency: The Trade-off Behind Your Bill
Serving more tokens per second and serving them faster are opposing goals. The knob that reconciles them is batch size, and it decides both speed and price.
Read
FundamentalsTime to First Token: What It Measures and What Moves It
TTFT is dominated by prefill compute, queueing and network distance rather than by model speed. What each contributes, and how to measure it without fooling yourself.
Read
GuidesTimeout Tuning for LLM Calls: Picking Numbers That Hold
Connect, first-token and idle timeouts do different jobs. How to pick each from your own latency data, propagate deadlines, and avoid the default ten-minute wait.
ReadShowing 589–600 of 731 articles