Model Performance

28 posts · 17 sources

+9

H2O.aiBlog5d

H2O-Lightning-4B: #1 on JevBench, and Ahead of Jev Itself

AI EngineerYouTube2w

Autoresearch Made Our Models 3x Faster — Tejas Bhakta, Morph

DeepMindBlog2w

Gemini 3.8 text-to-speech says hello

Jane StreetBlog4w

A study of sequence weighting at scale

AnthropicBlog4w

Measuring AI capabilities in intelligence targeting and conventional weapons

CohereBlog4w

Introducing North Small Translate: A leading sovereign open-weight machine translation model

Two Minute PapersYouTube1 Sept

GLM 5.3: Powerful AI Is Becoming Almost Free

Together AIBlog21 Aug

GLM-5.3 vs. Claude Fable 5 on DeepSWE: Cost, Coding, and Routing

Two Minute PapersYouTube19 Aug

DeepSeek Just Made Closed AI Look Ridiculous

Two Minute PapersYouTube5 Aug

The Billion Dollar AI Race Just Broke

VercelBlog27 Jul

DeepsecBench: evaluating model performance in finding cybersecurity vulnerabilities

BasetenBlog26 Jul

How we built the new fastest API for GLM-5.2

RampBlog17 Jun

You're Spending Too Much on AI. You're Also Using Too Little.

HarveyBlog28 May

Claude Opus 4.8, Now Live in Harvey

HexBlog21 May

Feature Engineering in Data Science: A Complete Guide

BasetenBlog3 Mar

Outperforming frontier models on emergency medicine documentation

HarveyBlog9 Feb

Expanding BigLaw Bench

BasetenBlog7 Aug

How we run GPT OSS 120B at 500+ tokens per second on NVIDIA GPUs

GoogleBlog22 Jul

Gemini 2.5 Flash-Lite is now stable and generally available- Google Developers Blog

Mistral AIBlog10 Jul

Upgrading agentic coding capabilities with the new Devstral models