KM4BI KM4BI

I Keep ChatGPT, Fable 5, and Grok 4.5 Open. Only One Gets My Hardest Work.

Nate Jones argues that model choice should follow your workflow, not a leaderboard: Soul, Fable, Grok, Luna and orchestration tools each fit different work…

Nate Jones makes a practical argument about model choice: the best model is not simply the one that tops a benchmark, but the one that helps you do your best work. His personal example is ChatGPT 5.6 Soul. He says it may not have the same “big model” feel for him, yet it fits his long, explicit prompting style and his structured knowledge-work tasks.

What the comparison shows

Soul is described as strong at reading long instructions, honoring explicit constraints and carrying complex work through to completion. Fable 5, by contrast, seems better suited to high-level intent, ambiguity, conceptual wrestling and front-end taste. For coding or agentic workflows, Nate also points to Luna, Grok, GLM 5.2 and Ringer, depending on cost, capability and orchestration needs.

The real selection criterion

The useful question is not “which model is best?” but “which model accelerates the loop that gets me to my best work?” Someone who dictates long, detailed prompts will need a different assistant from someone exploring unclear ideas or delegating coding tasks.

Why it matters now

Nate frames model families as having different characters. OpenAI’s 5.x line favors explicit direction and agentic execution; the Mythos/Fable lineage appears stronger at reading ambiguity and intent. He also highlights a gap: knowledge-work harnesses remain less mature than tools built for engineers, such as Codex or Claude Code.

Key takeaway

Test models on your own hard work. The model that makes you most comfortable doing that work is the one most worth keeping open.

Source