30
Perplexity Details Its GPU Embedding Stack: How Ivy, Tulip and ROSE Serve pplx-embed
Perplexity details its GPU embedding serving stack (Ivy, Tulip, ROSE), which reuses LLM prefill/decode kernels, CUDA graphs, and LazyTensors to cut launch overhead.
Perplexity engineers published a deep dive on the serving infrastructure behind pplx-embed, used across Perplexity Search and its API platform. The stack comprises Ivy (Rust HTTP gateway), Tulip (gRPC scheduling and batching), and ROSE (Runtime-Optimized Serving Engine), which reuses LLM prefill and decode kernels rather than running a separate embedding engine. Optimizations include whole-model CUDA graphs with lazy capture and a LazyTensor abstraction that overlaps CPU batch preparation with in-flight GPU work. Benchmarks are reported against vLLM v0.22.0 in BF16, with FlashAttention 4 generally fastest but FlashInfer 3 winning on Qwen-based models at very long sequence lengths.
33
So you want to use OpenRouter?
Simon Willison highlights OpenRouter pitfalls: provider routing makes the same model endpoint behave inconsistently across backends.
Simon Willison discusses problems, raised by Mohamed Moustafa, with relying on OpenRouter's single API endpoint that automatically routes to the most cost-effective provider. Different backend providers run different serving software with different optimizations and settings, so requests to the same endpoint can behave in different ways. Some providers lack vision capability for vision models, and the reasoning effort option can be processed differently across providers.
38
Implementation of Machine Learning Workflows with NVIDIA cuML, RAPIDS, GPU Benchmarking, Explainability, Clustering, and Model Inference
Hands-on tutorial implements NVIDIA cuML and RAPIDS to GPU-accelerate scikit-learn-style ML workflows with benchmarking, clustering, and inference.
The tutorial demonstrates NVIDIA cuML as a GPU-accelerated machine learning framework, using cuml.accel to speed up unmodified scikit-learn scripts with zero code changes and the native cuML API for CuPy/cuDF interoperability. It benchmarks CPU versus GPU implementations of PCA, K-Means, nearest-neighbor search, logistic regression, random forests, and DBSCAN on datasets up to 200,000 samples with 64 features. It also builds GPU pipelines with UMAP, t-SNE, and HDBSCAN, validates GPU-generated SHAP explanations, uses the FIL library for forest inference, and covers model serialization and GPU/CPU portability.
14
Speculative Decoding in vLLM on AMD GPUs
vLLM benchmarks speculative decoding on AMD Instinct MI300X and MI355X GPUs across five drafting methods including EAGLE-3 and native MTP.
The vLLM project documents draft-and-verify speculative decoding support for AMD GPUs via ROCm, comparing native MTP, Gemma 4 MTP, EAGLE-3, DFlash, and DSpark drafting approaches. Output-token throughput effects varied with drafting method, proposal length, model family, draft checkpoint, workload, and acceptance behavior. The post also covers how to enable each method plus practical tuning and observability considerations.
40
30
Show HN: Nari Qwen3-TTS and Qwen3-ASR – High accuracy, low latency and cost
Nari Labs claims top Coval voice AI benchmark rankings with low-latency, low-cost Qwen3-ASR and Qwen3-TTS inference endpoints.
Nari Labs says its Qwen3-ASR Fast endpoint ranks #1 in Coval's time-to-final-segment latency (p50 44 ms) with 3.6% WER at $0.12/hour, behind only AssemblyAI Universal 3.5 Pro on accuracy. Its Qwen3-TTS Fast ranks #2 in time-to-first-audio (p50 63 ms) and #1 in WER at 3.8%, priced at $10 per 1M characters. The company reports beating the official Qwen3 TTS Flash Realtime endpoint (8.8% WER, 692 ms median TTFA) and Baseten's dedicated endpoint (6.0% WER, 101 ms). Public beta APIs are moving to paid general availability with $20 in credits for existing accounts.
32
30
30
Agent-net Open Sources Webagent: A Go Harness That Turns Any Website into a Guarded AI Agent
Agent-net open-sourced Webagent, a Go harness turning websites into AI agents with code-enforced guardrails wrapping every tool call.
Agent-net released Webagent under Apache 2.0, a Go framework where a business fills in a declarative JSON spec, picks one provider for each of nine pluggable slots (retrieval, memory, guardrail, channel, secrets, presenter, model, action, observability), and runs webagent serve. Every tool the agent holds is wrapped by action.Guard so the chosen guardrail executes before any action runs and the model cannot bypass it. Live capabilities include OpenRouter/gateway LLM brains, MCP tools over Streamable HTTP, and Slack, WhatsApp, and HTTP channels; browser actions, OAuth-gated MCP, OTel export, and AgentNet identity/billing are not yet built. The project is v0 with a deferred-hardening list and cites arXiv 2511.19477 on an 85% versus 50% task-success gap attributed to architecture over model capability.
28
30
30
30
30
30
30
30
UC Berkeley Researchers Release CUA-Lite, an Open Platform Unifying Sandboxes, Data, Evaluation and RL for Computer-Use Agents
UC Berkeley's CUA-Lite is an open platform unifying computer-use agent sandboxes, datasets, evaluation and RL; Lite.OSWorld cuts OSWorld memory 4.1 GB to 0.9 GB.
UC Berkeley researchers released CUA-Lite, an open platform placing agents, environments, traces, and training for computer-use agents behind one action space, one LiteSample schema, and one command across desktop, browser, and mobile. Lite.OSWorld reproduces the OSWorld task suite and evaluators in plain Docker containers (0.9 GB RAM vs 4.1 GB, cold start 23.8s, ~4.6× more parallel instances), with scores matching the QEMU/KVM VM across 13 models. The platform claims 30k+ verifiable tasks, 15+ benchmarks, 10+ agents, and 20+ datasets on Hugging Face including Aguvis, OpenCUA, and ScaleCUA. A documented SFT run lifts Qwen3-VL-2B-Instruct mean episode return from 0.138 to 0.237 on the 332-task lite.osworld split.
45
30
30
30
30
30
30
30
AWS Introduces Pizza Bot: An Open Source Inbox for Background AI Agents
AWS open-sourced Pizza Bot, a self-hosted inbox app for background AI agents with approval gating and multi-provider model support.
AWS released Pizza Bot under Apache 2.0 after earlier versions served over 2,000 Amazon employees for meeting prep, email drafting, and research. The app provides macOS, Windows, and Linux desktop builds plus browser and terminal clients talking to a Hono API server, with LangGraph/DeepAgents checkpoints preserving thread state and approval pauses. It supports Amazon Bedrock, Anthropic, Google Gemini, OpenAI, OpenRouter, and Ollama, exposes external tools via MCP servers, and lets skill authors gate actions behind approve/edit/reject flows.
45
Traefik Labs brings independent verification to AI agent governance
Traefik Labs announces Sovereign Trust Plane in Traefik Hub, adding verifiable delegation, policy enforcement, and tamper-evident audit records for AI agent traffic.
Traefik Labs announced the Sovereign Trust Plane for Traefik Hub, generally available by September 30, 2026, providing delegated access, policy enforcement, and tamper-evident records for AI agent, tool, and API traffic. It implements the IETF ID-JAG draft with Okta Cross App Access and Janssen, enforces decisions through OpenID AuthZEN with OpenFGA and Cerbos, and commits cryptographic log fingerprints to transparency checkpoints verified by independently administered witnesses. The gateway also extends enforcement to MCP tool calls and the MCP server's backend API connection.
38
30
30
30
30
30
30
30
30
30
30
30
30
30