Models & Research

Market development: gpt-oss-120B

Hugging Face 2 months ago 6

ALTK-Evolve gpt-oss-120B

Read at Hugging Face

Agentic AI Model Optimization Agent Architecture

Benchmark result: OpenAI & GPT-5.6 Sol

Roboflow Blog 2 months ago 10 ● 3 sources

OpenAI GPT-5.6 Sol GPT-5.6 Terra GPT-5.6 Luna GPT-5.5 Gemini 3.5 Flash

Read at Roboflow Blog

AI Benchmarking

Market development · 18 Aug 2026

TheSequence 2 months ago 6

Read at TheSequence

Inference Optimization Advanced Reasoning Model Optimization

Acquisition: Google & Spirit Airlines

9to5Google 2 months ago 4 ● 5 sources

Google Spirit Airlines

Read at 9to5Google

Data Privacy Google Model Training

Model release: Cartesia & Sonic-3.6

MarkTechPost 2 months ago 6 ● 2 sources

Cartesia Sonic-3.6 Artificial Analysis ElevenLabs Eleven v3

Read at MarkTechPost

Model Release AI Benchmarking Text-to-Speech

Market development: Princeton & Claude Opus 4.8

MIT Technology Review 2 months ago 9

Princeton Claude Opus 4.8 NeurIPS

Read at MIT Technology Review

AI Agents Advanced Reasoning AI Capabilities

Market development: Exponential View

Exponential View 2 months ago 8

Exponential View

Read at Exponential View

AI & Academia

Market development: Tsinghua AIR & Seed1.6

MarkTechPost 2 months ago 2

ByteDance Seed Tsinghua AIR CUDA Agent Seed1.6 KernelBench PPO torch.compile SKILL.md

Read at MarkTechPost

AI Infrastructure Reinforcement Learning

Market development · 18 Aug 2026

Apple Machine Learning Research 2 months ago 0

Kernel-based optimal transport (OT) estimators Semismooth Newton method Short-step interior-point method (SSIPM) Vacher et al., 2021

Read at Apple Machine Learning Research

AI & Academia Scientific Research Model Optimization Machine Learning

Market development: MVICAD2

Apple Machine Learning Research 2 months ago 7

MVICAD2

Read at Apple Machine Learning Research

Scientific Research Multimodal AI

Market development · 18 Aug 2026

Apple Machine Learning Research 2 months ago 7

Group Relative Policy Optimization Reinforcement Learning with Verifiable Rewards

Read at Apple Machine Learning Research

Reinforcement Learning

Benchmark result: DeepSeek V4 Pro 0813 & GPT-5.6 Sol

Together AI 2 months ago 10 ● 2 sources

DeepSeek V4 Pro 0813 GPT-5.6 Sol DeepSWE DeepSeek OpenAI

Read at Together AI

AI Coding Agents Model Evaluation DeepSeek

Market development: Sentence Transformers & MultiVectorEncoder

Hugging Face 2 months ago 6

Sentence Transformers MultiVectorEncoder

Read at Hugging Face

Retrieval-Augmented Generation Embedding Models

Tuesday, 18 August 2026