Artificial Analysis now displays base model logos alongside derived video models on its leaderboards and has added an option to hide derived models to rank base models only.
Together has reached the top ranking for token volume across closed and open endpoints on OpenRouter, according to Vipul Ved Prakash.
Vidu has released Vidu Q4 Preview ahead of its full Q4 release, generating 3 to 16 second clips up to 4K with native audio and placing third on the Artificial Analysis AA-Video-I2V v1.0 leaderboard.
In an issue of The Batch, DeepLearning.AI analyzed DeepSeek's architecture, reporting an 890-byte cache per token that is 437 times smaller than DeepSeek-V1. The analysis also notes a 25% compute increase per output token from 4K to 1M input tokens and highlights Flash outperforming V4-Pro on the AA Index.
Polymarket reported that an Anthropic study found AI and robotics together are already capable of handling tasks that span 81% of U.S. employment.
Arena.ai evaluated typesafeai's Jev Router across more than 4,700 agentic sessions on Agent Arena, finding that while it does not improve on the current Pareto frontier, it demonstrates strong steerability in response to user feedback.
Hamed Nilforoshan evaluated OpenAI Decisions API against Jev on HiringCafe, reporting that OpenAI's API was twice as expensive and 5% to 10% worse for ranking job query relevance.
Sourcegraph demonstrated Claude Code querying code that has not been checked out locally by using Sourcegraph Deep Search. In an example task, the agent located deprecated packages across the Kubernetes codebase without exiting the Claude Code interface.
Grok Bot announced that it can now search, read, and monitor X.
Grok Bot is available to all users without requiring setup of the X connector. Users can direct the bot to track product feedback, follow breaking stories, or send weekly industry roundups.
Databricks has made Anthropic's Claude Haiku 5.5 available on its platform as a Day 0 launch, reporting roughly 15% higher quality than Haiku 4.5 on its OfficeQA Pro V1 benchmark.
Wispr Flow meeting notes can now be used as context in Cursor.
Artificial Analysis has launched AA-Music-Vocal v1.1 and AA-Music-Instrumental v1.1 for its Music Arena, evaluating models across 1,000 new prompts in 17 genres. Suno v6 ranks first on both leaderboards, followed by Suno v6-mini in second.
Together AI is scaling open-source model inference in collaboration with IBM and NVIDIA on a dedicated B300 cluster hosted on IBM Cloud.
OpenRouter has integrated Perplexity AI's open-weights multimodal decision model, Perplexity Decider v1.1 (pplx-decider-v1.1-27b). The model is priced at $0.02 per million input tokens with free output.
GitHub has made Anthropic's Claude Haiku 5.5 generally available across GitHub Copilot, including the Copilot app, CLI, and code. The model is designed for fast, high-volume tasks such as subagents and quick edits.
OpenRouter has made OpenAI's GPT-6 Luna Decisions available on its platform, featuring a 1M context window, pricing of $0.10 per million input tokens, and free output.
Command Code has introduced Claude Haiku 5.5 across its Pro and Max tiers as well as its API, offering a 1M token context window, up to 128k output tokens, and 75% lower costs than Haiku 4.5.
NVIDIA researchers developed PivotOPD, a method designed to teach AI agents how to avoid early task mistakes and recover when errors occur. The training approach uses a teacher model to show agents better actions and guide them back on track over subsequent steps.
AI startup BioinvestGPT forecasted the outcomes of five of six major drug trials before results were announced, according to Polymarket.
