Opinion: open-weight models can replace OpenAI and Anthropic
The author, a software and security engineer, argues that open‑weight large language models can serve as viable alternatives to OpenAI’s ChatGPT and Anthropic’s Claude. They cite GLM‑5.3 as roughly comparable to Claude Opus 5.5 for software engineering tasks and note that GLM‑5.3‑Flash approaches Claude Opus 4.8 while being smaller and cheaper. The piece stresses that these models can run…
Key points
- Open-weight models like GLM‑5.3 are claimed to be on par with Claude Opus 5.5 for software engineering
- GLM‑5.3 costs $1.80 per 1M input tokens, $0.80 per 1M cached tokens, $5.75 per 1M output tokens
- Tinfoil offers token‑based private inference; Maple AI provides a free plan and $20 / month Pro plan
Private‑inference services such as Tinfoil and Maple AI are highlighted: Tinfoil charges per token for enclave execution, while Maple offers a free tier and a $20 / month Pro plan for a ChatGPT‑like experience. The author also mentions OpenCode Go as a cheap, non‑private router costing $10 / month for credits. Pricing details for GLM‑5.3 are provided, and the writer concludes that open‑weight models can replace proprietary offerings at comparable or lower cost with fewer guardrails.
The article is an opinion piece, reflecting the author’s personal assessment and preferences rather than reporting verified corporate announcements.
Model pages: GLM 5.3 → · Claude Opus 5.5 → · Mistral Large 4 →
The story so far
5 episodes →- Opinion: open-weight models can replace OpenAI and Anthropicthis story
Just say no to OpenAI and Anthropic
micahflee.com · 12 October 2026
Loading the full article…
This text was published by micahflee.com and written by Micah Lee. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
1source- Hacker News discussion · 166 pointsnews.ycombinator.com
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
Comments
via GitHub Discussions