Qoder Adds Computer Use and Voice Control to Its Agentic Coding Platform
Qoder announced two new capabilities: Computer Use, which lets the AI see the screen, click, type and verify across desktop apps and browsers, and Qoder Voice,…
Qoder announced two new capabilities: Computer Use, which lets the AI see the screen, click, type and verify across desktop apps and browsers, and Qoder Voice,…
LiteLLM Lens now uses Typesafe's JEV to automatically judge each agent run and flag failures in logs, processing in about 130ms to handle 200K+ traces without m…
The opentunnel client package now ships Effect-based versions, letting developers embed end-to-end encrypted public URL tunneling into their apps via @opentunne…
A detailed article covering KV caching in flow models for image generation, including intuition, implementation details, benchmarks, and experimental results.
A benchmark comparing OpenAI's Decisions API against Jev on a real-world relevance-scoring task serving 2.5 million users found OpenAI to be 2x more expensive a…
Opentunnel offers public URLs for anything running on your machine with end-to-end encryption so the relay cannot see traffic, plus an SDK for embedding, set to…
Perplexity's open-weights multimodal decision model pplx-decider-v1.1-27b is now available on OpenRouter at $0.02/M input with free output, topping the Hugging…
LiteLLM announces day 0 support for Anthropic's Claude Haiku 5.5, which costs about a tenth of Haiku 4.5 per token, with detailed input/output/cache pricing tie…
OpenAI introduces Intelligent UI, giving ChatGPT visual, interactive answers, and begins a global rollout of GPT-6 (Sol for paid tiers, Luna for free tiers) acr…
Anthropic announces cache reads on Claude Sonnet 5.5 are being halved to $0.10 per million tokens, making it roughly 20% cheaper for most long-running work.