Tech Executive & AI Advisor
CTO / CIO / Consultant / Applied AI Practitioner
Explore Advisory Services →Bars: repositories with at least one commit that week — 85 at the busiest, out of 124 I build and run. 69% of commits paired with a coding agent. · commits
Bars are input tokens per week (incl. cache-creation/cache-read), left axis; the line is output tokens per week, right axis, independently scaled — not dollar cost. · output tokens
- Opus 5 -5.5pt 63.3%
- Sonnet 5 -7.5pt 27.9%
- Haiku 4.5 +5.5pt 4.4%
- Fable 5 -0.2pt 3.9%
- Opus 4.7 +0.7pt 0.2%
- Opus 4.8 new 0.2%
LLM Deliberation (multillm council) · 201,437 tokens, 08-17 – 09-07
- Gemini 3.1 Flash Lite 38,247 19%
- GPT-4.1 Mini 37,724 18.7%
- Llama 3.3 70B 36,844 18.3%
- DeepSeek V3 36,219 18%
- Gemini 3 Flash Preview (synth) 32,488 16.1%
- Mistral Small 3.1 19,915 9.9%
Long tail, tracked separately from the Claude mix above — % is a share of council/deliberation usage only, not the fleet total.