Launched this week on TrustList: Mercury 3

Inception's diffusion language model, released 19 October 2026 with a 1M-token context, at $0.20 per million input tokens; Inception says it runs at over 1,000 tokens per second.

#Mercury3 #AI #NewProduct #startups
trustlist.uk/launches/launch-in…

🌊 #InceptionLabs releases #Mercury3, the most capable diffusion-based #LLM so far. Benchmarks put it on par with or ahead of GPT-6 Luna and Claude Haiku 5.5 while running 4–8x faster #AI
🧵👇

Hi Folks! A hacker breaches 100+ firms for $50/hit w/ AI agents & leaves his dashboard open, the US bans #DeepSeek from agencies, & #Mercury3 writes 1K words/sec. +: #OpenAI translates 1950s bank code, #Anthropic tells the #IMF the singularity is here, & #Nature on AI consciousness. 🎧 #AI #TechNews

🧠 Inception, Mercury 3 modelini tanıttı!

Difüzyon mimarili model, şirketin testlerine göre NVIDIA GPU'larında saniyede 1.068 token üretebiliyor.
1 milyon token bağlam penceresi ve düşük maliyetli API erişimiyle özellikle geliştiricileri hedefliyor.

#Mercury3 #Inception #YapayZeka

Inception Releases Mercury 3, Its Flagship Diffusion LLM #Technology #AI #LargeLanguageModels #inception #mercury3 #llm #diffusionmodel #aitechnology

🚀 Inception just dropped Mercury 3, a diffusion LLM that cranks out 1,068 tokens/sec on NVIDIA GPUs. Bigger context window, lower latency—perfect for next‑gen voice bots. Curious how fast it really is? Dive in! #Mercury3 #DiffusionLLM #TokensPerSec

🔗 aidailypost.com/news/incepti...