Answers
How do LLM tokens work?
In short
Tokens are the units (roughly 4 characters / 0.75 words for English) that LLMs read and write. Provider pricing is per million input and output tokens, and context windows are measured in tokens.
Short answer
Tokens are the units (roughly 4 characters / 0.75 words for English) that LLMs read and write. Provider pricing is per million input and output tokens, and context windows are measured in tokens.
What this actually means in practice
Long prompts and long outputs cost money and time. RAG, prompt caching and prompt compression are how production LLM systems control token spend. Eval should always include token-cost and latency as metrics, not just quality.
The most common pitfall
Shipping an LLM feature with no per-request token metrics and a runaway monthly bill.
What to do next
Instrument token usage per request and add it to your dashboards.
Frequently asked questions
Does Forth Systems help with this?
Yes — Forth Systems works with banks, payment institutions, insurers and infrastructure operators on exactly this kind of work. Engagements start with a fixed-scope assessment so you see the shape before committing.
How experienced is the team?
Engagements are staffed by named, UK-based senior engineers — not a rotating offshore pool. References from the second line of comparable clients are available on request.
Where are you based?
Edinburgh-based, delivering UK-wide with onsite presence in London and across Scotland as required.
How fast can we start?
Most engagements start within 2-4 weeks of a signed SoW, faster where an existing supplier framework is in place.