Site-wide telemetry

The state of the board

Aggregate AI coding usage across every developer on Viberank — Claude Code, OpenAI Codex, Gemini CLI, Copilot and every agent ccusage tracks. Tracking 2024-07-23 → 2026-07-24.

Total spend

$6.7M

API-equivalent

Tokens

7.3T

all time

Developers

1.0K

on the board

Submissions

1.1K

tracked

Active days

564

distinct dates

Cache read

95%

of all tokens

Token breakdown

Input140.7B · 2%

Output25.6B · 0%

Cache read7.0T · 95%

Cache creation198.1B · 3%

95% of all tokens are cache reads — prompt caching is doing most of the heavy lifting.

Popular models

by users

claude-haiku-4-5651 devs

claude-sonnet-4-5451 devs

claude-sonnet-4378 devs

claude-sonnet-4-6375 devs

claude-opus-4-5341 devs

claude-opus-4-6315 devs

claude-opus-4-1243 devs

claude-opus-4-7237 devs

claude-opus-4233 devs

claude-opus-4-8230 devs

Top models

by spend

claude-opus-4-8$1.2M

gpt-5.5$926.4K

gpt-5.6-sol$904.4K

claude-fable-5$641.9K

claude-opus-4-7$299.5K

claude-opus-4-6$228.2K

gpt-5.4$129.4K

claude-sonnet-4-6$58.7K

claude-sonnet-5$41.0K

gpt-5.3-codex$40.2K

Tools

by users

Claude993 devs

Codex196 devs

Gemini76 devs

OpenCode70 devs

Other62 devs

Hermes44 devs

OpenClaw37 devs

pi17 devs

Kimi11 devs

Goose4 devs

Updated hourly. Spend is the API-equivalent cost reported by ccusage — most developers pay far less on subscription plans. Want on the board? npx viberank-cli