Grok Build with Grok 4.5 claims a 5× token-efficiency edge on coding tasks
elonmusk · x · 2026-07-27
Grok Build with Grok 4.5 is being presented as unusually token-efficient on coding-agent tasks. The post claims it sits alone in the chart’s most attractive efficiency quadrant, delivering top-tier coding-agent performance with about 3.5 million tokens per task.
By comparison, other agents in the same performance band reportedly use roughly 7 million to 18 million tokens per task—up to around 5× more. The pitch is straightforward: more work done within usage limits, faster execution, and better token economics for serious building. Tim Sweeney’s quote adds that it is very good for “vibe math” in programming language theory, and more fun to use than Claude Code-style workflows that overproduce docs or keep asking to continue.
More from coding & agent
- Inspired by OpenAI's 10,000-agent run, dev open-sources a crowdsourced agent problem-solving platform — Benjaminsen · 2026-09-11
- Lucid: open-source Mac app keeps your laptop awake only while AI agents run — Pitiful_Hedgehog_600 · 2026-09-11
- Alex Townsend posts 200 open problems in numerical linear algebra for humans and AI agents — IgorCarron · 2026-09-11
- Kimi K2.8 Preview rolls out: near-K3 coding performance, 1M context for all tiers — teortaxesTex · 2026-09-11
- Looking for a classifier of software engineering task shapes to pick models per task — StewartalsopIII · 2026-09-11
- Steal this idea: prompt-to-hardware where agents assemble custom devices — paraschopra · 2026-09-11