I Spent 400M Tokens Testing Kimi K3.1 — the Open-Source King Still Needs Work
I Spent 400M Tokens Testing Kimi K3.1 — the Open-Source King Still Needs Work Last weekend Kimi K3.1 Pro dropped, and honestly I was hyped. The last K3 I touche
9 نتيجة · “GPT-6”
OpenAI's strongest model — 1M context
“Most capable model I've used this year, period. The 1M context actually works for huge codebases”
OpenAI's cybersecurity reasoning model — #1 on ExploitBench, helped find 81 Firefox flaws and 948 Linux exploits. Gated to approved orgs (Daybreak Red).
“An AI that can discover real security vulnerabilities in the Linux kernel — this is the kind of progress that really matters.”
OpenAI's budget mid-tier model at half the price of Sol — fast on vision and computer use.
“AWESOME release for agentic production where cost matters — best API on the market”
OpenAI's professional-work model: financial analysis, legal research, slides and complex coding — 20% cheaper than GPT-6 Astra.
“Agents Last Exam +5.2 over GPT-6 Astra and +5.5 over Claude Fable 5.2 (OpenAI-reported).”
OpenAI's everyday-work model: coding, documents and multi-step tasks at about a fifth of Astra's per-task cost.
“About one-fifth of Astra's per-task cost.”
2.8T params, beats GPT-6 Astra on Terminal-Bench
“The 1M context actually works. I dumped a whole codebase in and it kept track of everything”
I Spent 400M Tokens Testing Kimi K3.1 — the Open-Source King Still Needs Work Last weekend Kimi K3.1 Pro dropped, and honestly I was hyped. The last K3 I touche
Model tested - Models: OpenAI GPT-6 Sol vs Anthropic Claude Opus - Use case: Coding, backend development, debugging, and AI agent workflows - Main evaluation cr
Codex right now feels dumb and lazy Compared to how blown away I was when 5.5 first dropped, 6 / 6.1 is honestly a big step down. The only time I get that old "