OpenCode Developer Pushes Back on Hidden Enterprise Demand for Open Source Models
An OpenCode developer argues that hidden enterprise demand for open source models is overstated, citing the small number of top inference providers and the scar…
An OpenCode developer argues that hidden enterprise demand for open source models is overstated, citing the small number of top inference providers and the scar…
Mistral introduced the Responses API in public preview for building agents, returning answers, reasoning and tool calls as separate items, and added DeepSeek V4…
An OpenCode developer says supply chain attacks are growing and that AUR must become more sophisticated or be replaced, arguing secure package publishing there…
Qoder changed the default reasoning setting for GLM-5.3, Kimi-K3, Kimi-K2.8-Preview and DeepSeek-Flash from Max to High, reporting a 40.5% drop in cost per run…
OpenAI's GPT-6.1 Terra is now available on B.AI's API and Web Chat, featuring a 1.05M context window, 128K max output tokens, adjustable reasoning effort, and p…
A Qwen ambassador built Qwenputer, which uses Qwen's Gated DeltaNet state as memory and embeds a 32-bit RISC-V CPU in the model weights, letting it execute C an…
B.AI's Model Leaderboard now ranks AI models across popularity, usage, price, and intelligence, adding weekly token share, multi-period API call stats, weighted…
MiniMax M3.1 Flash is available free on EvoX Agent from October 28 to 30, with the model described as an evolution of the Space Bunny version that topped global…
Mistral's Voxtral Mini Transcribe 3 placed second out of nine models by DER on Voice Arena's Diarization + ASR leaderboard, beating its larger counterpart.
ByteDance's Seedance video generation models are now available via API, including Seedance 2.5, 2.0, 2.0 Fast, and 2.0 Mini with different resolution, duration,…