All Stories — Page 11

The free Apache 2.0 Bonsai 2 models use ternary weights and add vision and tool calling, alongside performance figures reported by PrismML.

The October 15th update combines IBM's Granite models with task-specific adapters and a speech recognizer, aiming to run several workloads from one model deployment.

The company says the 23.5-square-kilometer reconstruction combines Lixel L2 Pro and PortalCam capture with an Unreal Engine workflow.

Nabiha Saklayen's consumer service stores cells below -150 degrees Celsius, while leaving any future medical use to the customer and their physician.

The release adds assistant- and terminal-based workflows; Kling says 10-bit HDR output and API access are still coming.

The feature pairs original music with effects placed at selected moments, and is available through ElevenCreative, ElevenMusic and ElevenAPI.

The investor wants agents that finish business workflows inside software companies already use, a problem his August research says depends on reliability beyond model scores.

The new showcase runs models including LingBot World and H3 Max Director on fal's WebRTC-based World Model Accelerator, whose developer API remains experimental.

The subscription bundles GLM 5.3 and Kimi K3 through Vercel AI Gateway, with zero data retention enabled by default and early access capped at 500 Pro teams.

The Singapore startup says its model ranked No. 2 in Artificial Analysis' silent image-to-video category and lists 1080p API generation at $0.12 per second.

A 2,500-task training set raised the model's score on AutomationBench-AA's public split from 34.0% to 57.7% in about 11 hours.

The six-model release pairs Arabic language models with speech recognition and publishes weights, training checkpoints and evaluation code under Apache 2.0.

The San Francisco startup says custom ROCm kernels and lower-precision math cut a 2K edit from about 22 seconds on the same MI355X hardware.

The test used simulated company workloads; Poetiq says its custom routers improved cost and accuracy against OpenRouter and TypeSafe's JEV on its benchmark.

Co-founded by Amazon-exit veteran Yogesh Ghaturle, Origin will build its OG-1 fleet and pilot the robot with contractors in the U.S. and Europe.

The October 15th update adds voices for French, German, Hindi, Mexican Spanish, Korean and Japanese; OpenAI says the API version will follow.

The product connects to observability and ticketing tools, investigates alerts and bugs, and can open tested pull requests; Cognition says it already handles thousands of alerts daily.

Comet 60B keeps text, image and video inputs, with downloadable weights in three formats; Multiverse's benchmark results are company-reported.

BODi confirmed she was one of 300,000 independent distributors in 2013; the company ended its multilevel marketing program in 2025.

Scott Reynolds told RuntimeWire the potential-cost estimates draw on customer and partner accounts, alongside third-party research.

CEO and co-founder María González Manso built Tucuvi around clinical follow-up; the program adds Microsoft sales channels and Marketplace procurement, not a reported customer deal.

FLUX 3 Image Klein adds editing and multi-reference generation, with weights for fine-tuning and an API priced from $0.009 per 1K images.

Isaiah Granet's latest product push automates website and API setup, while Bland's existing iMessage offering is still described as an enterprise service.

Better's first Better 2.0 board appointment comes days after shareholders restored founder Vishal Garg's control of the mortgage lender's board.