Cadence
Tokens burned per week 250B 200B 150B 100B 50B 0 Feb 2 Mar 2 Mar 30 Apr 27 Jun 1 partial
Monthly token burn per dev $100 $300 $1K $3K $10K monthly burn / dev developers →
GB of logs stored 4000 3000 2000 1000 0 Feb 5 Mar 5 Apr 2 Apr 30 Jun 3
Token burn per PR $1 $10 $100 $1K burn per merged PR merged PRs →
Median cost per line changed $1 $0.10 $0.01 $0.001 1-10 11-50 51-200 201-1k >1k lines changed $0.885 $0.205 $0.092 $0.022 $0.0028
% of sessions showing context rot
Claude Opus Codex
30% 25% 20% 15% 10% 5% 0% <0.5M 0.5-1M 1-2M 2-5M 5-10M 10-25M 25-100M 100M+ 23.2% 11.2%
Success rate: one-shot vs planning
Baseline One-shot
100% 75% 50% 25% 0% 89.1% 65.0% 81.6% 44.2% claude-code codex −24.1 pts −37.4 pts
Median initial context tokens
Claude Opus Codex
60k 50k 40k 30k 20k 10k 0 Feb 9 Mar 9 Apr 6 May 4 Jun 1 42k 17k
Skills + MCPs as % of initial context 18% 15% 12% 9% 6% 3% 0% Feb 9 Mar 9 Apr 6 May 4 Jun 1 ~15%
Loaded skill + MCP tokens that go unused 80% 60% 40% 20% 0% Feb 9 Mar 9 Apr 6 May 4 Jun 1 44%
Est. weekly tokens wasted on unused skills + MCP (log scale) 10B 1B 100M 10M 1M 100K Feb 9 Mar 9 Apr 6 May 4 Jun 1 partial wk
  1. 1Plan with the agent
  2. 2Review the plan in detail
  3. 3Agent implements
  4. 4Review the tests it built
  5. 5Push to CI → automated adversarial review
  6. 6Guided human review
Cadence Better AI coding for cheaper token bills.
teamcadence.ai Coding agent logs -> save money.
First month free CDNC-AI-ENG-MEL
001

Meta: Tokenmaxxxxxing

More adoption! Use the coding agent for everything.

  • Parallelism
  • Subagents
  • /goal
  • Dynamic workflows
  • Ralph loops
001

Meta: Tokenefficiency

Everyone's being moved off subscriptions.

  • Instantly 6x - 10x spend.
  • Orgs are budgeting $100k/dev/year for tokens.
  • Similar to cloud - devs control huge spend.
002

Real coding logs

We have terabytes of coding logs from our customers.
We do aggregated, anonymised analysis over them.

  • Vibe check.
  • Actionable insight about better AI coding.
003

Outcomes matter now

One imperfect metric is token burn per PR.

  • Bubble size is avg PR size.
  • Smart teams have relatively big PRs at relatively low cost.
  • High volume generally improves outcomes.
004

Smaller changes better?

Every session carries a big fixed cost: loading context, exploring, planning.

  • Smaller changes way more expensive per line.
  • But obviously easier to review and work with.
  • Action? find the sweet spot.
005

Bigger changes worse?

Measured using dev frustration rate and tool call failure rate.

  • Claude is worse as sessions get bigger.
  • Past 100m tokens it always gets worse.
  • But fine in "normal" usage ranges.
  • Action? find the sweet spot.
006

Plan vs one-shot

Everyone plans; I guess here's why.

  • Outcomes are much worse when one-shotting.
  • Action? planning is the job, get good at it.
007

Initial context is rising

System prompt, skills, MCPs.

  • Claude just keeps increasing.
  • Claude system prompt is much bigger.
008

Skills and MCPs

Skills are big.

  • Skills are install and forget.
009

Unused Skills or MCPs

Most aren't used.

010

Pure token wastage

Huge across our whole client base.

  • Action: time for a spring clean.
011

Simple recommended workflow

Balance up human and agent.

  • Review the plan.
  • Review the tests.
  • Guided review of the code and decisions.
012

Cadence

Visibility for eng leaders and coaching for devs.

  • Token and cost tracking across tools.
  • Per-developer tailored recommendations about levelling up.
  • PR reviews bringing together logs and code.

Hit me (Dave Slutzkin) up on LinkedIn for a chat, or use the coupon code.

↑ ↓ · space · or clicker to advance