GPT 5.6 Sol is really good and better than Claude Fable 5 in my opinion even on complex tasks.
It’s more token efficient, costs less per token, adding to that OpenAI provides better prompt caching.
It’s going to be my next workhorse.
Baptiste Jamin
@baptistejamin
Co-Founder/CEO @crisp_im - Speaking about AI, LLMs & code. - Bootstrapping & coding since 2015 the future of customer service.
- Note to the "AI experts" who ask us for a Model Router: A router that switches models mid-conversation breaks the cache, drops context, and costs MORE. The best thing it can do is not route. "But stateless / single-shot calls!" -> That's ~10% of your spend. You're adding a
- I appreciate @MistralAI, and I'm proud the EU has a company like this. As a founder pushing EU solutions at @Crisp, I genuinely want them to win. But how are crucial problems just... not getting fixed? Their TS SDK has 146 stars, and nobody seems to handle PRs or fix issues.
- Preparing a trip to canada: just searching for local food, picking the first result on Google. And seeing @crisp_chat 😆
- Again a massive supply chain attack on NPM. OIDC is solving absolutely nothing. The only solution is to ban using those pesky postinstall scripts, and eventually allow them for some packages.SECURITY ADVISORY — TanStack npm packages A supply-chain compromise affecting 42 @TansTack/* packages (84 versions total) was published to npm earlier today at approximately 19:20 and 19:26 UTC. Two malicious versions per package. Status: ACTIVE — packages are deprecated, npm
