Just for reference, from what I observed in my Using Local Coding Agents blog article last month: https://x.com/rasbt/status/207051816739969…

Just for reference, from what I observed in my Using Local Coding Agents blog article last month: https://x.com/rasbt/status/2070518167399698490?s=20 "I tried to analyze why Claude Code uses more tokens, and it seems that the difference mainly comes from input tokens rather than output tokens. In other words, Claude is not writing twice as much. The logs suggest that Claude is repeatedly feeding more context back into the model across turns, including previous messages, tool calls, command outputs, and file contents. For example, one Claude run used about 578k input tokens but only about 4.5k output tokens across 25 turns. So the likely explanation is that Claude’s harness accumulates or accounts for a larger prompt-side history during multi-step agent runs.” Have been taking different local open-weight LLMs for a test drive in different harnesses (Qwen-Code, Codex, Claude Code). 30B Mixture-of-Expert models are kind of a nice sweet spot and can solve challenging problems. And they get roughly 40 tok/sec on a Mac or DGX Spark, which i…

Related

• Consistent with what I found with Qwen3.6 a while back: Claude Code uses 2-3x as many tokens than (many) other harnesses at similar success …
• I put together a new article on setting up local coding agents with open-weight models. Everything runs 100% locally. I thought it might be …
• Link to the full article: https://magazine.sebastianraschka.com/p/using-local-coding-agents
• Using Local Coding Agents

Source: Sebastian Raschka (X) | 2026-07-30