A lot changed in the last quarter. New flagships from Anthropic, OpenAI, Google, and Meta. Open-weight models closing the gap. Reasoning models becoming table stakes. Cost dropping 20-50x since early 2024.

    I wrote up a concise survey of where things stand — frontier models, open-source contenders, and the trends that actually matter. No predictions. No "what this means for the industry." Just the models and what they changed.

    Read it here: https://proddraft.com/blog/state-of-llms-may-2026

    Next up: a practical comparison — running Llama 4 Maverick locally vs. calling Claude Sonnet 4 API for a real workload. Cost, quality, latency measured, not speculated.