Anthropic, OpenAI, Cloudflare and Google ship AI agents that find, verify and patch code bugs. How they compare.

It is the second 0.6B release in Edge0’s Audio8 TTS family. It builds on Audio8-TTS-Preview-0.6b and keeps the same ...

Prime Intellect’s Prime Mega MoE makes DeepSeek’s fused MoE kernel trainable, cutting Qwen3 training steps up to 55% ...

NVIDIA Releases Nemotron 3.5 Transcribe, an open speech LLM with 4.3B active parameters and a 4.71% English WER.

Samaya Sterling, a post-trained Nemotron 3.5 Lightning model with 3B active parameters, scores 58.9% on FrontierFinance at $0 ...

Gemini 4 Carbon, Google DeepMind’s frontier model, ranks #1 on Artificial Analysis Index at 59 and leads cybersecurity.

LEANN is a low-storage approximate nearest neighbor (ANN) index built on top of FAISS. It targets one problem: vector indexes ...

Block’s AI research lab has released Dessa, a full-duplex speech model trained only on human conversation. Block announced it on October 23, 2026 as a research preview. Dessa is a speech-to-speech ...

Liquid AI LFM3 ships 6 open-weight 2B-class models; LFM3-1.7B averages 49.37 across 29 self-reported benchmarks, above ...

AssemblyAI-TTS-1 is a streaming text-to-speech API for voice agents, priced at $15 per 1M characters with unlimited ...

OpenAI’s GPT-6.1 Terra brings fast computer use at $1/$5 per million tokens, scoring 60.0% on Agents’ Last Exam.

Google DeepMind’s Genie 4 world model generates interactive worlds with native audio, dialogue, and Gemini 3.8 Flash planning.