Desk archive
Cryptocurrency exchange Coinbase has begun integrating more affordable Chinese AI models like GLM 5.2 and Kimi 2.7, significantly reducing its AI spending while increasing token…
Read more
Sina Weibo's VibeThinker-3B, a compact 3-billion-parameter language model, demonstrates performance comparable to much larger models in math and coding tasks, suggesting that logical reasoning compresses…
Read more
Asian AI startups are rapidly developing and launching advanced AI models, including Mythos-like capabilities, as the US export ban on models like Anthropic's Mythos and…
Read more
OpenAI introduces GPT-5.6, a new family of tiered models including Sol, Terra, and Luna, designed to offer developers more flexibility in intelligence, speed, and cost.…
Read more
OpenAI has restricted the initial release of its new GPT-5.6 models, including flagship Sol, to a small group of partners after a request from the…
Read more
OpenAI has released its new flagship model, GPT-5.6 Sol, which outperforms Anthropic's Claude Mythos in coding benchmarks. However, its access is severely limited to select…
Read more
New data indicates Anthropic's Claude is increasingly attracting paid consumers, growing by 75% since January 2026, posing a challenge to ChatGPT's long-held market lead among…
Read more
DeepReinforce has released Ornith-1.0, an open-source family of AI models designed for agentic coding. These models distinguish themselves by learning to write their own execution…
Read more
A recent benchmark by Snowflake found Zhipu AI's GLM-5.2 model performing competitively against Anthropic's Claude Opus 4.7 in coding tasks, notably at a significantly lower…
Read more
NVIDIA's Canary-1B-v2 model offers developers a robust solution for multilingual automatic speech recognition (ASR) and speech translation, including automatic SRT subtitle export, using Python. This…
Read more