# Perplexity — How to Track and Control Team Token Spend

- Company: Perplexity (perplexity.ai)
- Announced: 2026-10-21
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://www.perplexity.ai/hub/blog/token-spend
- Record: https://forck.live/items/26949-how-to-track-and-control-team-token-spend
- Subject: Perplexity
- Pricing: In Perplexity's API pricing, some models charge six times as much for output

Perplexity's guide explains why team token spend grows and what teams can do to track and reduce it, aimed at people running AI on subscriptions or their own API builds. Token spend is the number of tokens a team consumes and what they cost, and tokens are not standardized, so the same text can produce different counts across tools. On subscription plans, token counts are rarely shown; many plans use credits instead, which a plan draws down as work gets done. In Perplexity, regular searches don't use credits, only Perplexity Computer does, and each task shows its own credit total in the conversation's Usage menu. Enterprise administrators can see how many people use Perplexity and how many queries they run weekly in the Insights dashboard, available to organizations with 50 or more seats or at least one Enterprise Max member. API providers show token counts directly: Perplexity's console tracks input, output, and reasoning token spend on the Billing page. Growth comes from agents running many steps, reasoning models writing out their thinking, and tokenmaxxing, where engineers run agents to raise personal token counts. To cut spend, Perplexity suggests starting new conversations when topics change, splitting long agent tasks, enabling context compaction, copying only relevant document sections, using model routing, and checking the first steps of agent tasks for loops.

## Evidence

Verbatim from https://www.perplexity.ai/hub/blog/token-spend:

> Every time a model replies, it usually rereads the whole conversation so far. Each new request includes all previous conversation history plus the current user message. So a short question late in a long thread uses more tokens than the same question in a new one.

---

Record: https://forck.live/items/26949-how-to-track-and-control-team-token-spend
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md