Welcome to LLMflation: LLM inference cost is going down fast
The single chart that explains why what is unaffordable today is cheap next year.
Open a16z.com →
Building & Product
Article
Thread
Article
Article
Article
Article
Article
Article
Guide
Article
Tool
Tool
Article
Article
Guide
Podcast
Article
Article
How much do AI agents cost to run, and how do I keep the bill under control?
The short answer
Token prices for equivalent quality fall roughly 10x a year, yet agent bills keep rising because agents burn 50 to 500 times the tokens of a chat: every step re-sends the whole history. A simple task can cost a fraction of a cent while a long agent session runs several dollars. The levers that matter: prompt caching (up to 90% off repeated context), routing easy steps to cheap models, trimming history, and measuring cost per completed task rather than per token.
A quick orientation. The real value is below: resources worth your time, from people who've actually done it.
18 resources worth your time
LLMflation: 10x cheaper per year
The headline numbers in thread form for a two-minute read.
Open x.com →AI Token Cost Over Time: Down 99.7%, Bills Up 3x
Names the paradox every founder hits: cheaper tokens, bigger bills.
Open navyaai.com →AI Agents Burn 50x More Tokens Than Chats
Explains the quadratic context growth that quietly multiplies agent costs.
Open leanopstech.com →Agentic AI Enterprise Token Cost
How large buyers model agent economics, useful when selling to them.
Open ey.com →AI Agent Cost: Token Economics & Optimization Guide
Concrete per-task cost ranges from cheap chatbot turns to $5+ agent runs.
Open cowork.ink →Token Cost Trap: Why Your AI Agent's ROI Breaks at Scale
Walks through the exact math of an agent whose unit economics collapse.
Open medium.com →Prompt caching with Claude
The feature that cuts up to 90% off repeated context, table stakes for agents.
Open anthropic.com →Prompt Caching with OpenAI, Anthropic, and Google Models
One page comparing how each provider's caching discount actually works.
Open prompthub.us →Context Engineering for AI Agents: Lessons from Building Manus
Why KV-cache hit rate is the one production metric that decides agent margins.
Open manus.im →AI Token Calculator
Estimate a task's cost across models before you commit to one.
Open tokencalculator.ai →AI Token Pricing Tracker
Live per-model pricing so your cost model never goes stale.
Open usagepricing.com →AI Token Costs: Why Enterprise AI Bills Keep Rising
Shows 72% of production cost sits outside the model invoice.
Open optimumpartners.com →Inference Unit Economics: The True Cost Per Million Tokens
The full stack of costs behind a token price, for founders selling AI products.
Open introl.com →Prompt Caching for Anthropic and OpenAI Models
A hands-on implementation guide with before-and-after cost numbers.
Open digitalocean.com →What Is an AI Agent? (pricing and economics discussion)
Investors debate agent pricing and unit economics in plain language.
Listen on Spotify open.spotify.com →Why 'Cheaper Tokens' Isn't a Good Reason to Own Your LLMs
Saves you from the premature self-host-a-model cost trap.
Open ml6.eu →AI's Plummeting Prices Are a Software Story, Not a Hardware One
Why the price curve should keep falling, and what that means for your roadmap.
Open weightythoughts.com →More in AI Agents
- What actually is an AI agent, in plain language? 18 →
- Why would a founder use agents instead of a chatbot or automation? 18 →
- What real tasks are founders delegating to agents today? 20 →
- How do I build my first agent without being technical? 20 →
- Which platform or framework should I build my agents on? 18 →
- How do I test an AI agent and know when it's reliable enough to trust? 18 →
- What can go wrong when an agent acts on my behalf, and how do I keep it safe? 16 →
- What is MCP, and why does it matter for my agents? 18 →
- Should I build one agent or a team of agents working together? 18 →
- Can coding agents like Claude Code actually build my product? 18 →
- Is there a real startup opportunity in building vertical AI agents? 19 →
- How should I price a product built on AI agents? 18 →
- Why do most AI agent projects fail, and what do the winners do differently? 19 →
- How do I give my agent the right context and memory so it stops forgetting? 18 →