Π£ΠΆΠ΅ Π΄ΠΎΡΡΡΠΏΠ½ΠΎ! ΠΡΡΠ»Π΅Π΄ΠΎΠ²Π°Π½ΠΈΠ΅ Telegram 2025 β ΠΊΠ»ΡΡΠ΅Π²ΡΠ΅ ΠΈΠ½ΡΠ°ΠΉΡΡ Π³ΠΎΠ΄Π° 
π¨ AI News | TestingCatalog
ΠΡΠΊΡΡΡΡ Π² Telegram
Latest AI News on AI Agents, Model Releases, Tools, Leaks, and Rumors π
ΠΠΎΠ»ΡΡΠ΅Π‘ΡΡΠ°Π½Π° Π½Π΅ ΡΠΊΠ°Π·Π°Π½Π°Π’Π΅Ρ
Π½ΠΎΠ»ΠΎΠ³ΠΈΠΈ ΠΈ ΠΏΡΠΈΠ»ΠΎΠΆΠ΅Π½ΠΈΡ16 268
6 578
ΠΠΎΠ΄ΠΏΠΈΡΡΠΈΠΊΠΈ
+2024 ΡΠ°ΡΠ°
+747 Π΄Π½Π΅ΠΉ
+44530 Π΄Π΅Π½Ρ
ΠΠ°Π³ΡΡΠ·ΠΊΠ° Π΄Π°Π½Π½ΡΡ ...
ΠΠΎΡ
ΠΎΠΆΠΈΠ΅ ΠΊΠ°Π½Π°Π»Ρ
5.7K
5.5K
5.2K
5.0K
4.2K
1
ΠΠ±Π»Π°ΠΊΠΎ ΡΠ΅Π³ΠΎΠ²
ΠΡ
ΠΎΠ΄ΡΡΠΈΠ΅ ΠΈ ΠΈΡΡ
ΠΎΠ΄ΡΡΠΈΠ΅ ΡΠΏΠΎΠΌΠΈΠ½Π°Π½ΠΈΡ
---
---
---
---
---
---
ΠΡΠΈΠ²Π»Π΅ΡΠ΅Π½ΠΈΠ΅ ΠΏΠΎΠ΄ΠΏΠΈΡΡΠΈΠΊΠΎΠ²
ΠΈΡΠ½Ρ '26
ΠΈΡΠ½Ρ '26
+90
Π² 1 ΠΊΠ°Π½Π°Π»Π°Ρ
ΠΌΠ°ΠΉ '26
+548
Π² 5 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°ΠΏΡΠ΅Π»Ρ '26
+666
Π² 5 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΡΡ '26
+495
Π² 6 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π²ΡΠ°Π»Ρ '26
+4 059
Π² 6 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ½Π²Π°ΡΡ '26
+603
Π² 1 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π΄Π΅ΠΊΠ°Π±ΡΡ '25
+308
Π² 2 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π½ΠΎΡΠ±ΡΡ '25
+472
Π² 6 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΎΠΊΡΡΠ±ΡΡ '25
+1 091
Π² 2 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π½ΡΡΠ±ΡΡ '25
+304
Π² 2 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°Π²Π³ΡΡΡ '25
+138
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΈΡΠ»Ρ '25
+232
Π² 1 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΈΡΠ½Ρ '25
+118
Π² 1 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΠΉ '25
+259
Π² 1 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°ΠΏΡΠ΅Π»Ρ '25
+15
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΡΡ '25
+25
Π² 2 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π²ΡΠ°Π»Ρ '25
+13
Π² 1 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ½Π²Π°ΡΡ '250
Π² 1 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π΄Π΅ΠΊΠ°Π±ΡΡ '24
+6
Π² 4 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π½ΠΎΡΠ±ΡΡ '24
+19
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΎΠΊΡΡΠ±ΡΡ '24
+49
Π² 2 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π½ΡΡΠ±ΡΡ '24
+24
Π² 1 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°Π²Π³ΡΡΡ '24
+25
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΈΡΠ»Ρ '240
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΈΡΠ½Ρ '24
+1
Π² 2 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΠΉ '24
+2
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°ΠΏΡΠ΅Π»Ρ '24
+172
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΡΡ '24
+4
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π²ΡΠ°Π»Ρ '24
+8
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ½Π²Π°ΡΡ '24
+5
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π΄Π΅ΠΊΠ°Π±ΡΡ '230
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π½ΠΎΡΠ±ΡΡ '230
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΎΠΊΡΡΠ±ΡΡ '230
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π½ΡΡΠ±ΡΡ '230
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°Π²Π³ΡΡΡ '23
+2
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΈΡΠ»Ρ '23
+2
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΈΡΠ½Ρ '23
+2
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΠΉ '230
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°ΠΏΡΠ΅Π»Ρ '230
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΡΡ '230
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π²ΡΠ°Π»Ρ '230
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ½Π²Π°ΡΡ '230
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π΄Π΅ΠΊΠ°Π±ΡΡ '220
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π½ΠΎΡΠ±ΡΡ '220
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΎΠΊΡΡΠ±ΡΡ '220
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π½ΡΡΠ±ΡΡ '220
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°Π²Π³ΡΡΡ '22
+10
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΈΡΠ»Ρ '22
+15
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΈΡΠ½Ρ '22
+20
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΠΉ '22
+11
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°ΠΏΡΠ΅Π»Ρ '22
+53
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΡΡ '22
+112
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π²ΡΠ°Π»Ρ '22
+27
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ½Π²Π°ΡΡ '22
+61
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π΄Π΅ΠΊΠ°Π±ΡΡ '21
+34
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π½ΠΎΡΠ±ΡΡ '21
+12
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΎΠΊΡΡΠ±ΡΡ '21
+13
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π½ΡΡΠ±ΡΡ '21
+37
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°Π²Π³ΡΡΡ '21
+258
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΈΡΠ»Ρ '21
+297
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΈΡΠ½Ρ '21
+74
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΠΉ '21
+187
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π°ΠΏΡΠ΅Π»Ρ '21
+149
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΠΌΠ°ΡΡ '21
+307
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ΅Π²ΡΠ°Π»Ρ '21
+603
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
ΡΠ½Π²Π°ΡΡ '21
+877
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
Get PRO
Π΄Π΅ΠΊΠ°Π±ΡΡ '20
+861
Π² 0 ΠΊΠ°Π½Π°Π»Π°Ρ
| ΠΠ°ΡΠ° | ΠΡΠΈΠ²Π»Π΅ΡΠ΅Π½ΠΈΠ΅ ΠΏΠΎΠ΄ΠΏΠΈΡΡΠΈΠΊΠΎΠ² | Π£ΠΏΠΎΠΌΠΈΠ½Π°Π½ΠΈΡ | ΠΠ°Π½Π°Π»Ρ | |
| 06 ΠΈΡΠ½Ρ | +13 | |||
| 05 ΠΈΡΠ½Ρ | +20 | |||
| 04 ΠΈΡΠ½Ρ | +18 | |||
| 03 ΠΈΡΠ½Ρ | +14 | |||
| 02 ΠΈΡΠ½Ρ | +12 | |||
| 01 ΠΈΡΠ½Ρ | +13 |
ΠΠΎΡΡΡ ΠΊΠ°Π½Π°Π»Π°
+1
BREAKING π₯: A new Claude Mythos 5 model slug has been spotted via Dev Mode.
Claude Mythos is planned to be released as its own model class, besides Haiku, Sonnet and Opus model families.
Soon? π
| 2 |  +1ANTHROPIC π₯: Claude Cowork limits have been doubled until July 5 for all paid plans.
Cowork time π | 702 |
| 3 |  +1GOOGLE π₯: NotebookLM will soon be able to generate files in many different formats from your sources, based on this teaser. There is a high chance that this release will be coupled with Gemini 3.5 Flash upgrade as well.
A huge list of formats referenced in the code.
["pdf","txt","md","docx","csv","pptx","epub","3g2","3gp","aac","aif","aifc","aiff","amr","au","avi","cda","m4a","mid","mp3","mp4","mpeg","ogg","opus","ra","ram","snd","wav","wma","avif","bmp","gif","ico","jp2","png","webp","tif","tiff","heic","heif","jpeg","jpg","jpe"] | 631 |
| 4 |  MICROSOFT π₯: Early look at Microsoft Scout Agent for Microsoft Frontier users.
Scout agent is designed for work use cases and was recently revealed at Microsoft Build 2026.
Features list π
> Models from OpenAI and Anthropic, including GPT-5.5 and Claude Opus 4.7
> Multi-step automations with support for browser use.
> Co-Create, an open canvas where users can collaborate with AI and export outputs as documents.
> Deep integration with Microsoft Teams
> Daily Briefing that summarises work context across connected services.
> Skills support, document generation, coding capabilities, and all that.
> Both macOS and Windows apps are available.
It would be quite a decent tool for Windows users! Yet, it is only for Frontier. Looking forward to testing their super app too. | 829 |
| 5 | Early Look: Microsoft rolls out Scout AI agent to Frontier users
Microsoft is rolling out Scout to Frontier organizations, previewing an always-on Autopilot agent for Microsoft 365. It runs on Windows and macOS, automates multi-step work, and operates under admin and tenant controls.
π #microsoftcopilot @testingcatalog | 768 |
| 6 |  MYTHOS π₯: Another early preview of recently spotted "Oceanus" checkpoint output.
"Oceanus" is rumored to be a version of the upcoming Mythos model, which is planned for public release within "weeks", according to Anthropic.
"Oceanus" prompt π | 868 |
| 7 | Anthropic started red teaming new Mythos models, first results
Anthropicβs Claude Oceanus preview points to a near-term release, centered on reasoning, coding, cybersecurity, and agentic tasks. Early reports suggest a sharp capability jump, as Anthropic also warns AI progress is accelerating AI research.
π #claude @testingcatalog | 800 |
| 8 | OpenSquilla lets AI agents organize their own skills
OpenSquilla shifts agent gains from bigger models to orchestration, using cost-aware routing and MetaSkill to discover and compose skills into validated workflows.
π #sponsored @testingcatalog | 814 |
| 9 | GOOGLE π₯: A new Troubleshooting mode has been spotted on Gemini.
In this mode, Gemini will explain troubleshooting process via text responses and interactive widgets. Even though it is working and available, it still looks like an unintended release and might get reverted soon.
Models for Troubleshooting π
h/t @AntiIkindaloveGemini | 1 043 |
| 10 |  NVIDIA π₯: Nemotron 3 Ultra has been released on Huggingface with 5x faster inference and 30% lower costs in comparison to other open models.
> Nemotron-3-Ultra-550B-A55B-NVFP4 is a frontier-scale large language model (LLM) trained by NVIDIA, designed to deliver strong agentic, reasoning, and conversational capabilities. | 1 009 |
| 11 |  OPENAI π₯: A new "more capable and scalable system for synthesizing memory" is being rolled out to Plus and Pro users in the US.
> Today, we are launching a significantly more capable and compute-efficient memory architecture built on top of dreaming.
> From the memory summary, you can quickly glean the highlights of what ChatGPT knows about you.
> This update will roll out to additional countries and Free and Go users over the coming weeks. | 948 |
| 12 |  ANTHROPIC π₯: A new internal research has been published, highlighting an accelerated AI development and a potential path to recursive self-improvement.
> Claude Mythos Preview could work for βat leastβ 16 hours and was βat the upper end of what [METR] can measure.β
> Today, Anthropic engineers on average ship 8x as much code per quarter as they did compared to 2021-2025.
Do you feel it? π | 955 |
| 13 |  ANTHROPIC π₯: A new "claude-oceanus-v1-p" has been made available to Red Teams.
This appearance may signal an upcoming release of newer Mythos models, referenced earlier by Antropic.
Soon? π | 985 |
| 14 |  Reve 2.0 is now available, and it landed in second place in the text-to-image arena, outranking Nano Banana 2.
> We invented a new way to generate and edit any image using precise layouts. For the first time, itβs possible to create images you can touch.
> Images are represented as code, so every part of an image becomes addressable, editable, and manipulable.
> Every image in Reve is segmented and labeled, giving you precise control over every region and element. | 1 135 |
| 15 | Microsoft Build 2026 recap, from Windows to Copilot, all AI
Microsoft used Build 2026 to push MAI as the in-house model stack for reasoning, coding, image, voice, and transcription. The key shift is strategic: less reliance on OpenAI, tighter integration across Copilot, Windows, and enterprise agents.
π #microsoftcopilot @testingcatalog | 1 033 |
| 16 |  GOOGLE π₯: A new Dreambeans experiment is now available in Google Labs for US-based Google AI Ultra users on the waitlist.
This experiment uses Personal Intelligence to deliver daily stories based on the user's data context.
Not a testing time for the most π | 1 095 |
| 17 |  Ideogram announced Ideogram 4.0, a new SOTA open image generation model.
> Ideogram 4.0 lands in the 8th spot on LM Arena and the 5th spot on Design Arena in the text-to-image category, and is getting close to Nano Banana Pro's performance.
> Ideogram 4.0 features dense, accurate text rendering, native 2K resolution, active background transparency, and precise layout control | 1 104 |
| 18 |  GOOGLE π₯: A new Gemma 4 12B is now available on Huggingface under Apache 2.0 license!
> Built with the same multimodal functionality as Gemma 4 E2B and E4B (text, audio, image, and video inputs), it brings native audio and vision understanding directly to local environments without the need for separate encoders.
> This unified approach to multimodality makes the model encoder-free, offering a deployment size that is perfect for consumer devices and streamlined local execution. | 1 242 |
| 19 |  Perplexity Personal Computer is now available to Max and Enterprise Max users on Windows!
Waitlist below π | 1 018 |
| 20 |  ICYMI π: Claude Code CLI can now operate Claude Platform, including the Messages API and Claude Managed Agents.
One CLI to rule them all π€ | 1 112 |
