Space Bunny Alpha is serving again on OpenRouter
OpenRouter says the issue that took Space Bunny Alpha offline is resolved and the model is serving again.
OpenRouter says the issue that took Space Bunny Alpha offline is resolved and the model is serving again.
On Artificial Analysis’s Pronunciation Robustness benchmark, Gemini 3.8 Flash TTS scores 89.5%, ahead of Gemini 3.1 Flash TTS, SpaceXAI TTS, and Gemini 3.8 Flash-Lite TTS. Leadership on individual pronunciation categories is split among several models.
Polymarket says the system is designed to run up to 380,000 sandboxes simultaneously and contain "agent misbehavior."
OpenAI says ChatGPT Voice can use plugins such as email, calendar, and Slack, run on GPT-6 Astra, Sol, and Luna, and work in ChatGPT Work on the web and mobile. The update is rolling out globally today in the latest version of the app.
Arena.ai reports that OpenAI’s GPT-6 Sol (Max) scored 1,689 points and placed fourth in Code Arena: WebDev, at a blended price of $8 per million tokens. It is 72 points above GPT-5.6 Sol (xHigh), while Arena.ai says GPT-6 Luna’s score is still pending.
OpenRouter lists Recraft V4.1 Flash at $0.007 per image and 1.3 seconds per image. It says the output still looks like Recraft for photography, portraits, complex scenes, logos, and early concepts.
Microsoft AI says MAI-Image-2.6 is available to try on Foundry and OpenRouter.
Microsoft AI puts that pace at 377 points a year, or 1.6 times the next-fastest climb on the board.
Microsoft AI says MAI gained 283 Elo on text-to-image in 10 months, putting its image-generation hill-climbing pace 61% ahead of any other lab while optimizing the Pareto frontier.
Polymarket says there is a 13% chance the United States enacts an AI safety bill. The associated market is https://poly.market/kJZeyQY.
According to Polymarket, the proposal would permanently ban superintelligent AI and impose prison sentences of up to 20 years for violations.
A schedule file in schedules/ that includes a cron expression and a prompt can be deployed so a Managed Deep Agent runs on its own.
The three models can be requested through one API from comfy, wavespeed, runware, fal, and higgsfield, with the destination chosen by price and availability.
GitHub will host beginner-friendly livestreams starting September 24 on building agentic applications with the GitHub Copilot SDK without implementing the agent loop. The series includes .NET and Java sessions, with sign-up in six languages.
Google Gemini can use WisprFlow to summarize meetings, capture voice notes, and extract action items, including decisions from the most recent meetings.
The example sends audio and telemetry from an ESP32-S3 to browsers over Cloudflare WebRTC, lets one listener control it, and is offered for people and their agents to build from.
The findings also say the approach can increase efficiency and support more advanced physical AI workloads.
Comfy Router keeps the same model string and arguments, with no new SDK, key, or redeploy. It uses Comfy credits, with no subscription or Router fee, and launches with Comfy, Runware, Wavespeed, Fal, and Higgsfield.
The new connections include adobe, squarespace, and onepeloton. Google Gemini says people can grow a business, design assets, and plan workouts in Gemini instead of switching between tabs.
Nimble says the plugin is available in the Plugins tab and lets ChatGPT and Codex combine web search with computer use, adjust retrieval after each run, and attach source controls, confidence scores, and evidence.