vibehacker
News
Anthropic ·

Claude Haiku 5.5 lands at $0.10/$0.50 per 1M tokens, 90% under Haiku 4.5 for prompts below 100K

Released Oct 7 for subagents and high-volume work, it scores 72.4% on OSWorld 2.1 (offline subset) and 39.2% on Terminal-Bench 4.0 vs 16.4% for GPT-6 Luna, and it's the first Haiku with adjustable effort (medium by default). Prompts over 100K tokens cost $0.50/$2.50, and since it uses slightly more tokens per text Anthropic puts the average saving near 75%; it's already live in GitHub Copilot and Amazon Bedrock.

Read at anthropic.com
Claude
AI assistant for research, coding, writing, and everyday work

More news

View all

Liquid AI open-sources d1-3B, a decision model that answers in 8 ms on an RTX 4090 with zero output tokens

Released Oct 7 on Hugging Face alongside the experimental d1 omni 600M (text plus image or audio), d1 3B returns a typed answer in one forward pass instead of generating text, scoring 48.57 on the Decision Index v0.2.1 public split, which Liquid says matches Decider 35B A3B at 12x smaller. It has day one llama.cpp support and runs a question in 16 to 50 ms on Jetson boards, so it's a cheap local option for gating, routing, or moderation steps in an agent loop…

Liquid AI Blog

Google ships an official agent skill and gcloud commands so coding agents read live Google docs instead of scraping

Announced Oct 7, npx skills add google/skills skill retrieving developer knowledge teaches Claude Code, Cursor, Copilot, or Antigravity to pull fresh Markdown docs for Google Cloud, Firebase, Android, and more through the Developer Knowledge MCP server, falling back to curl on the REST API. There's also a new gcloud developer knowledge surface (try piping an error trace into answer query ) plus client libraries in seven languages…

Google Developers Blog

Vercel puts stealth reasoning model Glyph Cluster free on AI Gateway for coding agents

Free during its stealth period for Pro and Enterprise teams with purchased AI Gateway credits, stealth/glyph cluster is a text only reasoning model for coding and long context work that you can pick in Claude Code, Codex, or Cursor after running vercel ai gateway setup . It has function calling but no structured outputs or image input, and there's no zero data retention, so your prompts and responses may be used for training…

Vercel Changelog

Copilot CLI 1.0.93 opens command sandboxing to all users and applies MCP config changes without a restart

Released Oct 7, anyone can now run shell commands in a sandbox with /sandbox or sandbox , MCP server config edits take effect between turns mid session, and enterprises get a permissions.limitTo setting that keeps network requests inside managed domains. Heads up: user settings are now read only from /.copilot/settings.json , so any user setting keys left in /.copilot/config.json are ignored…

GitHub / github/copilot-cli releases

Codex CLI 0.161.0 makes GPT-6.1 Sol the default model and adds `/mcp login` inside a running session

Released Oct 7, the update switches the bundled and Amazon Bedrock catalogs to GPT 6.1 Sol, lets you sign in to MCP servers mid session with /mcp login , and brings multi agent V2 and Ultra reasoning to Bedrock. Daybreak and automatic Cyber routing are now off unless you opt in with enable cli daybreak , and approved filesystem escalations can widen write access while keeping denied reads and network limits in place…

GitHub / openai/codex releases

Spotted something we missed? Start a thread.