Releases

Listen launched conversational Gabor-Granger pricing tests to measure willingness to pay.

Black Forest Labs released an open-weight 7B world-action model and reported first place on the RoboLab benchmark.

OpenEvidence announced a localized version of its clinical decision-support platform for healthcare providers across roughly 100 low- and middle-income countries.

OpenEvidence and Anthropic partnered to offer a localized clinical decision-support experience free to healthcare providers across roughly 100 low- and middle-income countries.

Skild published a post-training result showing a foundation policy learn dynamic, dexterous robotic behavior through long-horizon simulated self-play.

Harvey launched custom agents directly inside the Microsoft Word add-in.

Cursor launched Projects, a multi-agent environment with a coordinator capable of delegating work across very large numbers of subagents.

Cursor launched the Coordinator to plan and distribute complex project work among many specialized agents.

Palantir documented Model Context Protocol connectivity for governed access to Ontology data and actions.

AIP Evals added inline source-code support and related evaluation workflow improvements.

Anthropic introduced a new life-sciences research group and laboratory and reported that Claude agents identified a novel array-associated reverse-transcriptase system with CRISPR-like repeats.

The U.S. Department of Veterans Affairs selected Abridge to provide ambient AI under an enterprise contract.

Harvey expanded its legal research agent for end-to-end, citation-backed research workflows.

Databricks introduced Genie One MCP to provide trusted governed business context to external AI agents and cowork tools.

Scale and Google Cloud published a joint architecture for deploying Scale’s GenAI Portfolio on Google Cloud with Gemini Enterprise, identity, governance, and cloud infrastructure.

Transformers added support for running llama.cpp GGUF quantized models within the library’s model-definition and inference workflows.

Anthropic released Claude Opus 5.5, the first model in the Claude 5.5 family, with higher capability and lower operating cost than Opus 5.

OpenAI added GPT-6 Sol and GPT-6 Luna to the API as capability and efficiency tiers in the GPT-6 family.

Hugging Face released Tokenizers v1, a major milestone for its Rust-based tokenization stack.

Perplexity launched a marketplace for reusable Computer skills with enterprise governance.

Perplexity launched local-first Portable Computer for supported NVIDIA Windows and Linux systems.

Perplexity introduced configurable Light, Standard, High and Ultra effort levels for Computer tasks.

SpaceXAI released Grok 4.7 for coding and knowledge work with a larger base model, longer reinforcement-learning runs and stronger safeguards.

Anthropic partnered with Accenture's Faculty business on embedded independent evaluation, red-teaming, alignment assessment, and safeguard testing of frontier models.

fal reported H3 Max ranked first in its human-preference evaluation across quality, prompt understanding, and aesthetics, with top results also cited from independent benchmarks.