←Search results

Search “OpenAI”

By relevance · Total 304 results Clear search

Tuesday

Sunday

  2. Story6 reports · 1 facts

    Microsoft releases the Microsoft-Decision-1 decision scoring model

    On October 9, Microsoft launched the Microsoft-Decision-1 decision model. The model does not generate text; instead, it performs a single forward-pass scoring of preset answer options and returns a calibrated probability for each option, and can be used for routing, classification, prioritization, and verification. The model is post-trained on Alibaba's Qwen3.5-9B, is in public beta on Microsoft Foundry, and is available on OpenRouter under the identifier microsoft/microsoft-decision-1; its weights are not distributed publicly. Microsoft says it achieved 83.5% accuracy on 36 benchmarks and about 150,000 held-out questions, and plans to later migrate the base to MAI and OpenAI models. The OpenRouter listing shows a price of $0.042 per million input tokens and free output. Other reports say it is 2.5 times faster than the second-place H2O-Lightning-4B.

    5-9B, is in public beta on Microsoft Foundry, and is available on OpenRouter under the identifier microsoft/microsoft-decision-1; its weights are not distributed publicly. Microsoft says it achieved 83.5% accuracy on 36 benchmarks and about 150,000 held-out questions, and plans to later migrate the base to MAI and OpenAI models. The OpenRouter listing shows a price of $0.042 per million input tokens and free output. Other reports say it is 2.5 times faster than the second-place H2O-Lightning-4B.

    Matched fieldsSummary

Thursday

Wednesday

  1. Story4 reports · 1 facts

    Microsoft releases preview of image editing model MAI Image 2.7

    On October 27, 2026, Microsoft AI released the image editing model MAI Image 2.7, launching it in preview form. The Microsoft AI Blog said that the model ranks 3rd on Arena Edit and 2nd on the Artificial Analysis image editing leaderboard, making Microsoft AI the 2nd-ranked lab; the company also said that it outperforms 2.6 on all 10 use cases and 7 categories of editing actions evaluated by Artificial Analysis, and leads the next-best model by 24 Elo on scenes and styles. A subsequent IT Home report said that the model ranks 2nd on the AA-Image-Editing v2.1 leaderboard with 1193 Elo, second only to OpenAI's GPT Image 2.5 Sunburst (max). Users can currently try it on Arena, Artificial Analysis, and the MAI Playground, with text-to-image functionality coming soon; Microsoft has not yet announced a public API, pricing, or an official commercial launch date, and the model weights have not been released.

    s, and leads the next-best model by 24 Elo on scenes and styles. A subsequent IT Home report said that the model ranks 2nd on the AA-Image-Editing v2.1 leaderboard with 1193 Elo, second only to OpenAI's GPT Image 2.5 Sunburst (max). Users can currently try it on Arena, Artificial Analysis, and the MAI Playground, with text-to-image functionality coming soon; Microsoft has not yet announced a public API, pricing, or an official commercial launch date, and the model weights have not been released.

    Matched fieldsSummary

Monday

  1. Story1 reports · 1 facts

    Canva's first self-developed image model ranks second on DesignArena

    Canva has launched Canva Craft, its first self-developed frontier image model, and it ranks second on the DesignArena image generation leaderboard, behind only OpenAI. The model consists of two parts: a planner and a renderer. The planner first writes out a composition plan, and the renderer then draws it at up to 2K resolution and in more than 30 aspect ratios. Because the layout and text are determined before rendering, editing a single element changes only that element. There is currently no information on further developments.

    Canva has launched Canva Craft, its first self-developed frontier image model, and it ranks second on the DesignArena image generation leaderboard, behind only OpenAI. The model consists of two parts: a planner and a renderer. The planner first writes out a composition plan, and the renderer then draws it at up to 2K resolution and in more than 30 aspect ratios. Because the layout and text are determined before rendering, editing a single element changes only that element. There is currently no

    Matched fieldsSummary

Wednesday

  2. Story4 reports · 1 facts

    ElevenLabs releases the Eleven Realtime v1 speech-to-speech model

    On October 20, 2026, ElevenLabs released the Eleven Realtime series of speech-to-speech models, comprising v1 Mini for low-cost natural conversation and v1 Max for complex tasks. According to the ElevenLabs blog, both support more than 90 languages and can access its over 17,000 voices; an independent evaluation by Artificial Analysis shows that v1 Max answered 98.8% of Big Bench Audio speech reasoning questions correctly and began speaking in an average of 0.57 seconds, while v1 Mini scored 75.9% and 0.62 seconds, respectively. The API is compatible with the OpenAI Realtime API, so existing integrations need only change the endpoint and model name, and models can also be selected directly for deployment in ElevenAgents. On October 22, Zhidongxi reported that the two models currently support only WebSocket connections; in its testing, an assistant built with Architect could modify Agent settings by voice, but the Chinese accent was noticeable, responses tended to be long, and latency was relatively high.

    io speech reasoning questions correctly and began speaking in an average of 0.57 seconds, while v1 Mini scored 75.9% and 0.62 seconds, respectively. The API is compatible with the OpenAI Realtime API, so existing integrations need only change the endpoint and model name, and models can also be selected directly for deployment in ElevenAgents. On October 22, Zhidongxi reported that the two models currently support only WebSocket connections; in its testing, an assistant built with Architect could

    Matched fieldsSummary

Friday

Thursday

Monday

Tuesday

Thursday