Holo Company launches Holo3, SOTA Computer Use model
Holo Company has introduced Holo3, an enterprise automation model scoring 78.85% on OSWorld-Verified while operating at one-tenth the cost of larger rivals. With open weights, API access, and agentic training, it supports complex, multi-application business workflows.
π #ai
Holo Company has introduced Holo3, an enterprise automation model scoring 78.85% on OSWorld-Verified while operating at one-tenth the cost of larger rivals. With open weights, API access, and agentic training, it supports complex, multi-application business workflows.
π #ai
TestingCatalog
Holo Company launches Holo3, SOTA Computer Use model
What's new? Holo Company launched Holo3 for business automation scoring 78.85% on OSWorld; the Holo3-35B-A3B weights are public under Apache2 with a free API tier;
β€2π2
Media is too big
VIEW IN TELEGRAM
Slackbot is now an always-on agent that leverages business knowledge from Slack context.
New stuff π
- Meeting intelligence - Slackbot now listens in the background of your meetings.
- Slackbot now travels across your desktop as you work and understands the context of whatever you're doing.
- Reusable AI skills! π₯
- Slackbot can now route work to Agentforce or any agent or app in your enterprise.
- Native CRM for small teams.
- Slack as the conversational interface for Salesforce.
- Deep research and voice input! π₯
New stuff π
- Meeting intelligence - Slackbot now listens in the background of your meetings.
- Slackbot now travels across your desktop as you work and understands the context of whatever you're doing.
- Reusable AI skills! π₯
- Slackbot can now route work to Agentforce or any agent or app in your enterprise.
- Native CRM for small teams.
- Slack as the conversational interface for Salesforce.
- Deep research and voice input! π₯
π6π₯1
Google prepares 3D Avatars for Gemini and Remy tools for learning
Recent Gemini builds reveal an Avatar option for creating 3D portraits, a student-focused Remy learning mode, and signs of Skill Support reaching some Ultra users.
π #gemini
Recent Gemini builds reveal an Avatar option for creating 3D portraits, a student-focused Remy learning mode, and signs of Skill Support reaching some Ultra users.
π #gemini
TestingCatalog
Google prepares 3D Avatars for Gemini and Remy tools for learning
Google is testing a 3D Avatar option in Geminiβs attachment menu and a Remy learning mode, as signs point to more student features, potentially before Google I/O.
β€5π2
Anthropic tests new Claude Code desktop UI amid source code leak
Anthropic appears to be testing Epitaxy, a possible new interface for Claude Code Desktop featuring expanded hotkeys and playful visuals, as a source map leak exposed internal models like Capybara and Strudel, fueling speculation about its roadmap.
π #claude
Anthropic appears to be testing Epitaxy, a possible new interface for Claude Code Desktop featuring expanded hotkeys and playful visuals, as a source map leak exposed internal models like Capybara and Strudel, fueling speculation about its roadmap.
π #claude
TestingCatalog
Anthropic tests new Claude Code desktop UI amid source code leak
Anthropic is testing a new Claude Code mode called Epitaxy, featuring extra hotkeys and playful visuals, and source code leaks are sparking user interest.
β€4π2 2
BREAKING π¨: Gemma 4 is likely to be released by Google on Thursday!
A good day to test local models π
A good day to test local models π
β€6π4
Exclusive: Anthropic tests its own always-on "Conway" agent
Anthropic is developing Conway, a standalone Claude environment with extensions, webhooks, Chrome connectivity, and always-on agent capabilities.
π #claude
Anthropic is developing Conway, a standalone Claude environment with extensions, webhooks, Chrome connectivity, and always-on agent capabilities.
π #claude
TestingCatalog
Exclusive: Anthropic tests its own always-on "Conway" agent
Anthropic is testing Conway, a standalone Claude agent environment featuring extensions, webhooks, and Chrome use, hinting at always-on support.
β€5π1
π¨ AI News | TestingCatalog
Exclusive: Anthropic tests its own always-on "Conway" agent Anthropic is developing Conway, a standalone Claude environment with extensions, webhooks, Chrome connectivity, and always-on agent capabilities. π #claude
This media is not supported in your browser
VIEW IN TELEGRAM
BREAKING π¨: ANTHROPIC IS WORKING ON ITS OWN ALWAYS-ON AGENT SOLUTION CALLED CONWAY!
CONWAY WILL HAVE A SEPARATE UI INSTANCE, WILL BE ABLE TO OPERATE BROWSER, CONNECTORS, CLAUDE CODE (EPITAXY?) AND COULD BE INVOKED VIA WEBHOOKS.
IT WILL ALSO SUPPORT EXTENSIONS, AN UPCOMING CNW ZIP STANDARD FROM ANTHROPIC TO BUILD CUSTOM TOOLS, UI TABS, AND CONTEXT HANDLERS!
LET CLAUDE COOK π₯
CONWAY WILL HAVE A SEPARATE UI INSTANCE, WILL BE ABLE TO OPERATE BROWSER, CONNECTORS, CLAUDE CODE (EPITAXY?) AND COULD BE INVOKED VIA WEBHOOKS.
IT WILL ALSO SUPPORT EXTENSIONS, AN UPCOMING CNW ZIP STANDARD FROM ANTHROPIC TO BUILD CUSTOM TOOLS, UI TABS, AND CONTEXT HANDLERS!
LET CLAUDE COOK π₯
π₯8 1 1
Meta tests Paricado model family, Health, and Document agents
Hands-on testing indicates Metaβs Avocado models are further along than reported, with multiple variants, including a reasoning build, already partially active on the background. A new Paricado family spotted as well.
π #meta
Hands-on testing indicates Metaβs Avocado models are further along than reported, with multiple variants, including a reasoning build, already partially active on the background. A new Paricado family spotted as well.
π #meta
TestingCatalog
Meta tests Paricado model family, Health, and Document agents
Meta is testing new Parikado AI models, Health and Document agents. Previously spotted Avocado variants being A/B tested.
BREAKING π¨: Google released Gemma 4 in 4 different variants: 31B, 26B MoE, 2B and 4B!
Offline on-device AI apps got a huge upgrade!
Offline on-device AI apps got a huge upgrade!
π₯11β€5π1
Media is too big
VIEW IN TELEGRAM
Microsoft now has MAI-Transcribe-1, MAI-Voice-1, and MAI-Image-2 available on MAI Playground!
- "MAI-Voice-1 is our top-tier voice generation model. Built to generate natural, realistic speech, rich with nuance, emotional range, and expression that preserves speaker identity even across long-form content."
- "MAI-Transcribe-1 delivers state-of-the-art speech-to-text transcription across the top 25 most-used languages according to the industry-standard FLEURS benchmark."
- "MAI-Image-2 has turbocharged image generation performance and speed on Copilot after debuting as a top 3 model family on the Arena leaderboard."
- "MAI-Voice-1 is our top-tier voice generation model. Built to generate natural, realistic speech, rich with nuance, emotional range, and expression that preserves speaker identity even across long-form content."
- "MAI-Transcribe-1 delivers state-of-the-art speech-to-text transcription across the top 25 most-used languages according to the industry-standard FLEURS benchmark."
- "MAI-Image-2 has turbocharged image generation performance and speed on Copilot after debuting as a top 3 model family on the Arena leaderboard."
π6
ChatGPT voice mode can now be used in CarPlay on iOS 26.4+
Needs testing π
Needs testing π
π5π₯2
BREAKING π¨: OpenAI aquired TBPN daily live show!
βItβs one of the places where the conversation about AI and builders is actually happening day to day. A lot of you already watch it, and rely on it to stay close to whatβs going on.β
βItβs one of the places where the conversation about AI and builders is actually happening day to day. A lot of you already watch it, and rely on it to stay close to whatβs going on.β
β€5π3π€2
Media is too big
VIEW IN TELEGRAM
Microsoft released Copilot Advisors experiment in the US where multiple AI characters with live portraits will debate on a certain topic.
Users will be able to select from various personalities and specify any topic they want. At the end, the debate will be summarized on a page.
Copilot war room π
Users will be able to select from various personalities and specify any topic they want. At the end, the debate will be summarized on a page.
Copilot war room π
π₯5π2β€1