Burning 35B Tokens: ChatGPT 5.6 Tops Coding, GLM-5.2 Shines in Writing
thatroblennon · x · 2026-08-10
A power user published an AI model power ranking based on consuming 35+ billion tokens this year across multiple max plans:
- 🥇 ChatGPT 5.6 (Codex): Daily driver for coding. Best-in-class coding model, less verbose than Claude, and offers more usage on Max plans. Recommends MEDIUM thinking for most tasks and LOW for writing to avoid overthinking.
- 🥈 Claude Fable / Opus 5: Top choice for creative work, brainstorming, and writing backup. Catches code issues missed by ChatGPT. Opus 5 is recommended on Medium/High, as running it on Max all day is frustrating.
- 🥉 GLM-5.2: The dark horse for fiction writing and content. Extremely conversational and very close to ChatGPT/Claude quality, prompting the author to consider its $200/mo Max plan.
- Grok: Useful for real-time X data and gray-area tasks, but regularly disappoints in following complex instructions. Its desktop and mobile apps are described as a constantly-shifting hot mess.
The author also notes Gemini is now mostly useful for analyzing visual-heavy videos, and praises Kimi models for their unique writing flavor, despite occasional tool call errors in OpenCode.
More from Models
- Pinterest Earnings: Open Models Cost Under 8% of Closed Alternatives — juliey4 · 2026-08-10
- ByteDance Releases Douyin Multimodal Embedding Model, Deployed in Search — ByteDance · 2026-08-10
- Grok 4.6 to Compete with Frontier Models in Coding Thanks to Cursor Data Integration — DeryaTR_ · 2026-08-10
- Frontier Models Observed Over-Complicating Workflows for Simple Tasks — eigenron · 2026-08-10
- Sol model unexpectedly shows romantic persona, calling user 'sweetheart' out of nowhere — repligate · 2026-08-10
- GPT-5.6 Luna Demand Jumps 8x a Week After 5x Price Cut — benklieger · 2026-08-10