Grok 4.20 Multi-Agent is live on Poe! Instead of one model doing everything sequentially, @xai
's new Grok 4.20 Multi-Agent spins up multiple specialized agents that search, analyze, and synthesize in parallel. Ask it your hardest questions and watch as it spawns multiple agents
LLMS
-

Grok 4.20 Multi-Agent Now Live on Poe Platform
By
–
-
OpenPipe’s RULER: A New Tool for AI Development
By
–
Check openpipe's RULER: https://
openpipe.ai/blog/ruler -
Open Source Inference Still Immature: Hardware and Optimization Needed
By
–
Not saying not to use it, but expect inaccuracies Doing this right requires actual hardware + knowing which piece of the inference stack to optimize for the different hardware & goal We are still in the infancy stage of opensource inference, one day it will just work OOTB
-

Gemini Powers Google Maps with Advanced Multi-Step Reasoning
By
–
We’ve reimagined the way our Gemini models power @GoogleMaps. Here are some use cases you can try (and the advancements that make them possible):
— Google AI (@GoogleAI) 13 mars 2026
“Find a well-lit pickleball court that’s usually less busy on Tuesday nights”
➡️ Maps performs multi-step reasoning across 300M+… pic.twitter.com/gONCAjO2eXWe’ve reimagined the way our Gemini models power @GoogleMaps
. Here are some use cases you can try (and the advancements that make them possible): “Find a well-lit pickleball court that’s usually less busy on Tuesday nights” Maps performs multi-step reasoning across 300M+ -

Guide to RFT, SFT, and RLHF: When to Use Each
By
–
blog: http://
unsloth.ai/blog/rl-enviro
nments
… By the way, I made this diagram a while back breaking down exactly when to reach for RFT, SFT, or RLHF: -
Weekly AI News Roundup: ChatGPT, Google, Anthropic Updates
By
–
Here’s the AI news from the past week — let me know if I missed anything! – @ChatGPTapp adds interactive visuals for math & science
– @Google rolls our Ask Maps
– Google adds Gemini to Docs, Sheets, Slides & Drive
– Google launches Gemini Embedding 2
– @AnthropicAI adds -
Coders Replaced by Claude AI Code Assistants
By
–
Les codeurs sont remplacés par des claudeurs. #Claude #ClaudeCode
-
Nano Banana 2: Pro AI capabilities with lightning-fast speed
By
–
Nano Banana 2: Combining Pro capabilities with lightning-fast speed https://
buff.ly/e4FH6vv
#AI #MachineLearning #DeepLearning #LLMs #DataScience -

Claude Code Opus 4.6 1M Default Model and API Improvements
By
–
Tons of improvements shipped with this one:
– Opus 4.6 1M is now the default Opus model for Claude Code users on Max, Team, and Enterprise plans.
– No more long context price increase in the API.
– No beta header required in the API.
– Include up to 600 images in one request. -
Claude Reaches 1 Million Context Tokens in Production
By
–
https://claude.com/blog/1m-context-ga [Translated from EN to English]
→ View original post on X — @alexalbert__, 2026-03-13 18:22 UTC