Loading activity…
@togethercompute
Accelerate inference, model shaping, and pre-training on a research-optimized platform.
AI
Together says developers can define training loops using reinforcement learning or supervised fine-tuning while it handles the infrastructure.
AI
finkd says the model runs well on new Mac Studios, crediting work with Apple and John Ternus.
Technology
NVIDIA says the model combines reasoning and tool use with text, image, document and video understanding.
YouTubeAI
Nvidia says the open model combines reasoning and tool use with text, image, document and video understanding.
YouTubeAI
NVIDIA says Step 5 Preview runs at nearly 8K tokens per second per GPU on GB300 using NVFP4.
AI
Together AI says its NYC gathering with NVIDIA and Reflection drew a full room of builders.
AI
The organizer says founders, researchers and builders gathered with Nvidia and DTCP Capital; former Blue Angels also appeared.
AI
Together AI says enterprises can now run open models at production scale on the dedicated IBM Cloud cluster.
AI
n8n says a member of its team will talk about building with open frontier models alongside Nvidia.
AI
Together’s Auto router selects a model once per session; an optional Anthropic key changes the choices.
AI
Together Compute says the lineup covered how to shortlist open models and evaluate them against production traffic, not just benchmarks.
AI
Together said it ranked No. 1 in OpenRouter token share on September 29, 2026, for GLM 5.3 Flash (29.2%), DeepSeek V4.1 Flash (25.6%) and Kimi K3 (18.9%).
AI
Together AI pitched Alibaba Qwen’s model for high-volume uses such as coding and coworking assistants, and suggested running evaluations during the offer.
AI
Nvidia says its open-source OpenShell runtime traces agent actions and enforces policies. Sentry adds monitoring on BlueField hardware, designed to enable millisecond-scale quarantine.
AI
Together shares a quote describing frontier models as “really strong” but “really slow for real-time use cases,” alongside a look at its work with Decagon.
AI
A commenter questions how a prediction could be disproved without knowing what people might have privately. The reply limits the comparison to what is publicly available.
AI
Together AI announced a session with speakers from Ollama and Together AI on the infrastructure needed to build coding agents and run them at scale.
AI
Together AI says teams working at scale discussed cost, model selection, routing and reliability at its London gathering.
AI
Together AI says startup delays recur on every model turn in agent workflows, making lower time to first token more important as those loops grow longer.
AI
Together AI says the video model generates 4–15-second clips at up to 2K resolution with native stereo audio.
