𝗠𝗲𝗱𝗶𝗮𝗧𝗲𝗸 𝗤𝟯’𝟮𝟲: 𝗥𝗲𝘃𝗲𝗻𝘂𝗲 𝗚𝗿𝗲𝘄 𝟭𝟭% 𝗬𝗼𝗬, 𝗘𝘅𝗰𝗲𝗲𝗱𝗲𝗱 𝗛𝗶𝗴𝗵 𝗘𝗻𝗱 𝗼𝗳 𝗚𝘂𝗶𝗱𝗮𝗻𝗰𝗲; 𝗦𝗺𝗮𝗿𝘁 𝗘𝗱𝗴𝗲 𝗢𝗳𝗳𝘀𝗲𝘁 𝗠𝗼𝗯𝗶𝗹𝗲 𝗣𝗵𝗼𝗻𝗲 𝗗𝗲𝗰𝗹𝗶𝗻𝗲 @MediaTek'𝗌 𝖰𝟥 𝟤𝟢𝟤𝟨 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝖾𝗑𝖼𝖾𝖾𝖽𝖾𝖽 𝗍𝗁𝖾 𝗁𝗂𝗀𝗁 𝖾𝗇𝖽 𝗈𝖿 𝗂𝗍𝗌 𝗀𝗎𝗂𝖽𝖺𝗇𝖼𝖾, 𝗋𝖾𝖺𝖼𝗁𝗂𝗇𝗀 𝖴𝖲$𝟧.𝟥 𝖻𝗂𝗅𝗅𝗂𝗈𝗇, 𝗎𝗉 𝟣𝟣% 𝖸𝗈𝖸. 𝖬𝗈𝖻𝗂𝗅𝖾 𝖯𝗁𝗈𝗇𝖾 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝖽𝖾𝖼𝗅𝗂𝗇𝖾𝖽 𝟤𝟥% 𝖸𝗈𝖸, 𝖺𝗅𝗍𝗁𝗈𝗎𝗀𝗁 𝗍𝗁𝗂𝗌 𝗐𝖺𝗌 𝗉𝖺𝗋𝗍𝗅𝗒 𝗈𝖿𝖿𝗌𝖾𝗍 𝖻𝗒 𝗀𝗋𝗈𝗐𝗍𝗁 𝗂𝗇 𝗍𝗁𝖾 𝖲𝗆𝖺𝗋𝗍 𝖤𝖽𝗀𝖾 𝗌𝖾𝗀𝗆𝖾𝗇𝗍. 𝗞𝗲𝘆 𝗧𝗮𝗸𝗲𝗮𝘄𝗮𝘆𝘀 • 𝖪𝖾𝗒 𝗀𝗋𝗈𝗐𝗍𝗁 𝖽𝗋𝗂𝗏𝖾𝗋𝗌 𝖽𝗎𝗋𝗂𝗇𝗀 𝗍𝗁𝖾 𝗊𝗎𝖺𝗋𝗍𝖾𝗋 𝗂𝗇𝖼𝗅𝗎𝖽𝖾𝖽 𝖼𝗈𝗇𝗍𝗂𝗇𝗎𝖾𝖽 𝗌𝗍𝗋𝖾𝗇𝗀𝗍𝗁 𝗂𝗇 𝖲𝗆𝖺𝗋𝗍 𝖤𝖽𝗀𝖾 𝗉𝗅𝖺𝗍𝖿𝗈𝗋𝗆𝗌 𝖺𝗇𝖽 𝗍𝗁𝖾 𝗋𝖺𝗆𝗉-𝗎𝗉 𝗈𝖿 𝖠𝖲𝖨𝖢 𝗉𝗋𝗈𝖽𝗎𝖼𝗍𝗂𝗈𝗇, 𝗐𝗁𝗂𝖼𝗁 𝗉𝖺𝗋𝗍𝗅𝗒 𝗈𝖿𝖿𝗌𝖾𝗍 𝗐𝖾𝖺𝗄𝗇𝖾𝗌𝗌 𝗂𝗇 𝖬𝗈𝖻𝗂𝗅𝖾 𝖯𝗁𝗈𝗇𝖾 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝖺𝗆𝗂𝖽 𝗋𝗂𝗌𝗂𝗇𝗀 𝗆𝖾𝗆𝗈𝗋𝗒 𝗉𝗋𝗂𝖼𝖾𝗌 𝖺𝗇𝖽 𝗌𝗈𝖿𝗍𝖾𝗋 𝖽𝖾𝗆𝖺𝗇𝖽. • 𝖲𝗆𝖺𝗋𝗍 𝖤𝖽𝗀𝖾 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝗀𝗋𝖾𝗐 𝟧𝟥% 𝖸𝗈𝖸, 𝖽𝗋𝗂𝗏𝖾𝗇 𝖻𝗒 𝗏𝗈𝗅𝗎𝗆𝖾 𝗉𝗋𝗈𝖽𝗎𝖼𝗍𝗂𝗈𝗇 𝖺𝖼𝗋𝗈𝗌𝗌 𝟧𝖦 𝗆𝗈𝖽𝖾𝗆𝗌, 𝖠𝖨 𝖯𝖢𝗌, 𝗍𝖺𝖻𝗅𝖾𝗍𝗌 𝖺𝗇𝖽 𝖺𝗎𝗍𝗈𝗆𝗈𝗍𝗂𝗏𝖾, 𝖺𝗅𝗈𝗇𝗀𝗌𝗂𝖽𝖾 𝗁𝗂𝗀𝗁𝖾𝗋 𝖳𝖵 𝖲𝗈𝖢 𝗋𝖾𝗏𝖾𝗇𝗎𝖾. • 𝖬𝗈𝖻𝗂𝗅𝖾 𝖯𝗁𝗈𝗇𝖾 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝖽𝖾𝖼𝗅𝗂𝗇𝖾𝖽 𝟤𝟥% 𝖸𝗈𝖸 𝗂𝗇 𝖰𝟥, 𝗂𝗆𝗉𝖺𝖼𝗍𝖾𝖽 𝖻𝗒 𝗆𝖾𝗆𝗈𝗋𝗒 𝗌𝗎𝗉𝗉𝗅𝗒 𝖼𝗈𝗇𝗌𝗍𝗋𝖺𝗂𝗇𝗍𝗌, 𝗁𝗂𝗀𝗁𝖾𝗋 𝖽𝖾𝗏𝗂𝖼𝖾 𝖼𝗈𝗌𝗍𝗌 𝖺𝗇𝖽 𝗌𝗈𝖿𝗍𝖾𝗋 𝖾𝗇𝖽-𝗆𝖺𝗋𝗄𝖾𝗍 𝖽𝖾𝗆𝖺𝗇𝖽, 𝗉𝖺𝗋𝗍𝗂𝖼𝗎𝗅𝖺𝗋𝗅𝗒 𝗂𝗇 𝗍𝗁𝖾 𝗆𝗂𝖽- 𝖺𝗇𝖽 𝗅𝗈𝗐-𝖾𝗇𝖽 𝗌𝖾𝗀𝗆𝖾𝗇𝗍𝗌. • 𝖯𝗈𝗐𝖾𝗋 𝖨𝖢 𝖼𝗈𝗇𝗍𝗋𝗂𝖻𝗎𝗍𝖾𝖽 𝟧% 𝗈𝖿 𝗍𝗈𝗍𝖺𝗅 𝗋𝖾𝗏𝖾𝗇𝗎𝖾, 𝗀𝗋𝗈𝗐𝗂𝗇𝗀 𝟣𝟪% 𝖸𝗈𝖸, 𝖽𝗋𝗂𝗏𝖾𝗇 𝖻𝗒 𝗆𝖺𝗋𝗄𝖾𝗍-𝗌𝗁𝖺𝗋𝖾 𝗀𝖺𝗂𝗇𝗌 𝗂𝗇 𝖽𝖺𝗍𝖺 𝖼𝖾𝗇𝗍𝗋𝖾-𝗋𝖾𝗅𝖺𝗍𝖾𝖽 𝖻𝗎𝗌𝗂𝗇𝖾𝗌𝗌. • 𝖠𝖨 𝖠𝖲𝖨𝖢 𝗌𝗁𝗂𝗉𝗆𝖾𝗇𝗍𝗌 𝗌𝗍𝖺𝗋𝗍𝖾𝖽 𝗂𝗇 𝖲𝖾𝗉𝗍𝖾𝗆𝖻𝖾𝗋, 𝗐𝗂𝗍𝗁 𝗌𝗍𝗋𝗈𝗇𝗀 𝖼𝗎𝗌𝗍𝗈𝗆𝖾𝗋 𝖽𝖾𝗆𝖺𝗇𝖽 𝗌𝗎𝗉𝗉𝗈𝗋𝗍𝗂𝗇𝗀 𝖾𝗑𝗉𝖾𝖼𝗍𝖾𝖽 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝗈𝖿 𝖴𝖲$𝟤.𝟧 𝖻𝗂𝗅𝗅𝗂𝗈𝗇 𝗂𝗇 𝖰𝟦 𝟤𝟢𝟤𝟨.   • 𝖰𝟦 𝟤𝟢𝟤𝟨 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝗂𝗌 𝖾𝗑𝗉𝖾𝖼𝗍𝖾𝖽 𝗍𝗈 𝖻𝖾 𝗂𝗇 𝗍𝗁𝖾 𝗋𝖺𝗇𝗀𝖾 𝗈𝖿 𝖴𝖲$𝟨.𝟨–𝟨.𝟫 𝖻𝗂𝗅𝗅𝗂𝗈𝗇, 𝗋𝖾𝗉𝗋𝖾𝗌𝖾𝗇𝗍𝗂𝗇𝗀 𝟥𝟩–𝟦𝟥% 𝖸𝗈𝖸 𝗀𝗋𝗈𝗐𝗍𝗁, 𝗉𝗋𝗂𝗆𝖺𝗋𝗂𝗅𝗒 𝖽𝗋𝗂𝗏𝖾𝗇 𝖻𝗒 𝗋𝖾𝗏𝖾𝗇𝗎𝖾 𝗋𝖾𝖼𝗈𝗀𝗇𝗂𝗍𝗂𝗈𝗇 𝖿𝗋𝗈𝗆 𝗍𝗁𝖾 𝖠𝖲𝖨𝖢 𝖻𝗎𝗌𝗂𝗇𝖾𝗌𝗌 𝖺𝗇𝖽 𝖼𝗈𝗇𝗍𝗂𝗇𝗎𝖾𝖽 𝗀𝗋𝗈𝗐𝗍𝗁 𝗆𝗈𝗆𝖾𝗇𝗍𝗎𝗆 𝗂𝗇 𝗍𝗁𝖾 𝖲𝗆𝖺𝗋𝗍 𝖤𝖽𝗀𝖾 𝗌𝖾𝗀𝗆𝖾𝗇𝗍. 𝗥𝗲𝗹𝗮𝘁𝗲𝗱 𝗜𝗻𝘀𝗶𝗴𝗵𝘁𝘀 𝗠𝗲𝗱𝗶𝗮𝗧𝗲𝗸 𝗗𝗶𝗺𝗲𝗻𝘀𝗶𝘁𝘆 𝟵𝟲𝟬𝟬 𝗣𝗿𝗼: 𝗠𝗼𝘃𝗶𝗻𝗴 𝘁𝗼 𝟮𝗻𝗺 𝘄𝗶𝘁𝗵 𝗮 𝗙𝗼𝗰𝘂𝘀 𝗼𝗻 𝗣𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗮𝗻𝗱 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆: 𝗠𝗲𝗱𝗶𝗮𝗧𝗲𝗸 𝘁𝗼 𝗖𝗼𝗻𝘁𝗿𝗶𝗯𝘂𝘁𝗲 𝟭 𝗶𝗻 𝟰 𝗔𝗜 𝗔𝗦𝗜𝗖 𝗦𝗲𝗿𝘃𝗲𝗿 𝗖𝗼𝗺𝗽𝘂𝘁𝗲 𝗦𝗵𝗶𝗽𝗺𝗲𝗻𝘁𝘀 𝗶𝗻 𝟮𝟬𝟮𝟴: 𝗗𝗮𝘁𝗮 𝗖𝗲𝗻𝘁𝗲𝗿 𝗢𝗽𝗽𝗼𝗿𝘁𝘂𝗻𝗶𝘁𝗶𝗲𝘀 𝗔𝗰𝗰𝗲𝗹𝗲𝗿𝗮𝘁𝗲 𝗠𝗲𝗱𝗶𝗮𝗧𝗲𝗸’𝘀 𝗗𝗶𝘃𝗲𝗿𝘀𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝘆 #MediaTek# #Earnings# #Semiconductors# #SmartEdge# #AIASIC# #ASIC# #EdgeAI# #MobileSoC#

.@GSMA @M360series LATAM 2026 is underway in Mexico City (May 13–14). Our Chief International Ecosystem Representative, Summer Chen, joins @SpringhamJustin of @mobileworldlive to discuss how #ZTE# is driving #AI# value for operators across LATAM. Watch the full interview. 👇 #M360#

Mobile Coin (MOB) will be listed on Bitfinex! #MOB# deposits are now open. Trading and withdrawals for @mobilecoin on #Bitfinex# will start from 02/04/21 at 09:00 AM UTC. Discover $MOB available pairs⬇️

𝗟𝗲𝘁'𝘀 𝗖𝗼𝗻𝗻𝗲𝗰𝘁 𝗮𝘁 𝗠𝗩𝗡𝗢 𝗡𝗮𝘁𝗶𝗼𝗻 𝗘𝘂𝗿𝗼𝗽𝗲 𝟮𝟬𝟮𝟲 𝖳𝗁𝖾 𝖾𝗏𝗈𝗅𝗎𝗍𝗂𝗈𝗇 𝗈𝖿 𝖾𝖲𝖨𝖬 𝗂𝗌 𝖼𝗋𝖾𝖺𝗍𝗂𝗇𝗀 𝗇𝖾𝗐 𝗈𝗉𝗉𝗈𝗋𝗍𝗎𝗇𝗂𝗍𝗂𝖾𝗌 𝖿𝗈𝗋 𝖬𝖵𝖭𝖮𝗌 𝗂𝗇 𝖨𝗈𝖳 𝖺𝗇𝖽 𝖾𝗇𝗍𝖾𝗋𝗉𝗋𝗂𝗌𝖾 𝖼𝗈𝗇𝗇𝖾𝖼𝗍𝗂𝗏𝗂𝗍𝗒. 𝖡𝗎𝗍 𝗁𝗈𝗐 𝖼𝖺𝗇 𝗍𝗁𝖾 𝗂𝗇𝖽𝗎𝗌𝗍𝗋𝗒 𝗍𝗋𝖺𝗇𝗌𝗅𝖺𝗍𝖾 𝗍𝗁𝖾 𝖾𝗆𝖾𝗋𝗀𝗂𝗇𝗀 𝖲𝖦𝖯.𝟥𝟤 𝗌𝗍𝖺𝗇𝖽𝖺𝗋𝖽 𝗂𝗇𝗍𝗈 𝗋𝖾𝖺𝗅-𝗐𝗈𝗋𝗅𝖽 𝖽𝖾𝗉𝗅𝗈𝗒𝗆𝖾𝗇𝗍𝗌 𝖺𝗇𝖽 𝗇𝖾𝗐 𝖻𝗎𝗌𝗂𝗇𝖾𝗌𝗌 𝗈𝗉𝗉𝗈𝗋𝗍𝗎𝗇𝗂𝗍𝗂𝖾𝗌? 𝖩𝗈𝗂𝗇 𝗈𝗎𝗋 𝖱𝖾𝗌𝖾𝖺𝗋𝖼𝗁 𝖣𝗂𝗋𝖾𝖼𝗍𝗈𝗋, Mohit Agrawal (@MohitOpinion), 𝖺𝗍 MVNO Nation 𝖤𝗎𝗋𝗈𝗉𝖾 𝟤𝟢𝟤𝟨, 𝗐𝗁𝖾𝗋𝖾 𝗁𝖾 𝗐𝗂𝗅𝗅 𝗌𝗉𝖾𝖺𝗄 𝗈𝗇 “𝗡𝗲𝘄 𝗲𝗦𝗜𝗠 𝗦𝘁𝗮𝗻𝗱𝗮𝗿𝗱 𝗳𝗼𝗿 𝗜𝗼𝗧 – 𝗛𝗼𝘄 𝗖𝗮𝗻 𝗠𝗩𝗡𝗢𝘀 𝗦𝗲𝗶𝘇𝗲 𝘁𝗵𝗲 𝗢𝗽𝗽𝗼𝗿𝘁𝘂𝗻𝗶𝘁𝘆?” 𝖧𝖾 𝗐𝗂𝗅𝗅 𝖾𝗑𝗉𝗅𝗈𝗋𝖾 𝖲𝖦𝖯.𝟥𝟤 𝖺𝖽𝗈𝗉𝗍𝗂𝗈𝗇, 𝖾𝖺𝗋𝗅𝗒 𝖽𝖾𝗉𝗅𝗈𝗒𝗆𝖾𝗇𝗍 𝗂𝗇𝗌𝗂𝗀𝗁𝗍𝗌, 𝖺𝗇𝖽 𝗍𝗁𝖾 𝗈𝗉𝗉𝗈𝗋𝗍𝗎𝗇𝗂𝗍𝗂𝖾𝗌 𝗍𝗁𝗂𝗌 𝗇𝖾𝗐 𝗌𝗍𝖺𝗇𝖽𝖺𝗋𝖽 𝗉𝗋𝖾𝗌𝖾𝗇𝗍𝗌 𝖿𝗈𝗋 𝖬𝖵𝖭𝖮𝗌. 𝗗𝗮𝘁𝗲: 𝖭𝗈𝗏𝖾𝗆𝖻𝖾𝗋 𝟦, 𝟤𝟢𝟤𝟨 𝗧𝗶𝗺𝗲: 𝟫:𝟣𝟧 – 𝟫:𝟥𝟢 𝖠𝖬 𝗣𝗹𝗮𝗰𝗲: 𝖧𝗈𝗍𝖾𝗅 𝖬𝖾𝗅𝗂á 𝖵𝗂𝗅𝗅𝖺𝗂𝗍𝖺𝗇𝖺, 𝖠𝗅𝗂𝖼𝖺𝗇𝗍𝖾, 𝖲𝗉𝖺𝗂𝗇 𝖫𝖾𝗍'𝗌 𝖼𝗈𝗇𝗇𝖾𝖼𝗍 𝖺𝗇𝖽 𝖾𝗑𝗉𝗅𝗈𝗋𝖾 𝗍𝗁𝖾 𝖿𝗎𝗍𝗎𝗋𝖾 𝗈𝖿 𝖨𝗈𝖳 𝖼𝗈𝗇𝗇𝖾𝖼𝗍𝗂𝗏𝗂𝗍𝗒! 𝖤𝗆𝖺𝗂𝗅 𝗂𝗇𝖿𝗈@𝖼𝗈𝗎𝗇𝗍𝖾𝗋𝗉𝗈𝗂𝗇𝗍𝗋𝖾𝗌𝖾𝖺𝗋𝖼𝗁.𝖼𝗈𝗆 𝗍𝗈 𝗌𝖼𝗁𝖾𝖽𝗎𝗅𝖾 𝖺 𝗆𝖾𝖾𝗍𝗂𝗇𝗀 𝗐𝗂𝗍𝗁 𝖬𝗈𝗁𝗂𝗍. 𝗥𝗲𝗮𝗱 𝗼𝘂𝗿 𝗹𝗮𝘁𝗲𝘀𝘁 𝗿𝗲𝗽𝗼𝗿𝘁 𝘁𝗼 𝘂𝗻𝗱𝗲𝗿𝘀𝘁𝗮𝗻𝗱 𝗵𝗼𝘄 𝗲𝗦𝗜𝗠 𝗼𝗿𝗰𝗵𝗲𝘀𝘁𝗿𝗮𝘁𝗶𝗼𝗻 𝗶𝘀 𝗲𝘃𝗼𝗹𝘃𝗶𝗻𝗴 𝗮𝗻𝗱 𝘂𝗻𝗹𝗼𝗰𝗸𝗶𝗻𝗴 𝗻𝗲𝘄 𝗴𝗿𝗼𝘄𝘁𝗵 𝗼𝗽𝗽𝗼𝗿𝘁𝘂𝗻𝗶𝘁𝗶𝗲𝘀: #MVNONationEurope# #MVNO# #eSIM# #SGP32# #IoT# #EnterpriseConnectivity# #MobileConnectivity#

We release Needle 3: A Sliceable 8-29MB automation foundation model that can match DeepSeek V4 Flash. One set of weights, every depth from 2 to 20 layers a model of its own, 25-121M parameters at CQ2-bit, built on our Simple Attention Networks and running locally at up to 4k tokens/sec decode speed on a Raspberry Pi 5. Needle does not chat. Every turn is a function call: give it the tools your app exposes and it picks the right ones and fills every argument from what the user said, or hand it a schema and it returns a typed record. Ask for something no tool covers and you get an empty list, not a guess. That trade is lets 121M parameters trained on 360B tokens of structured data beat models 10x their size on mobile tool calls and match 2-3x bigger models on structured JSON extraction. It runs on mobiles, wearables, smart home devices, small robots and microcontrollers, with prebuilt engines for macOS, Linux, Windows, Android, iOS, watchOS, tvOS, the browser and WASI hosts. Try it in your browser:

🤖 Mobile GUI agents are fine on short tasks but fall apart on long, multi-step ones. The reason is simple: just piling up action and screenshot history blows up the context and dilutes the cross-app information that actually matters. This work's idea is to promote context management from "passively accumulating history" to "an action the agent deliberately chooses." Just as it picks a UI operation, the agent can also choose to fold its history and UI state. It keeps three structured fields — folded action history, folded UI state, and a recent step record — balancing the overall arc with the immediate details. That's the Context-as-Action framework, MemGUI-Agent. 🎯 It's trained on MemGUI-3K (2,956 annotated trajectories), producing the 8B MemGUI-8B-SFT via supervised fine-tuning. The result reaches best open-data 8B performance on MemGUI-Bench and generalizes to the out-of-distribution MobileWorld benchmark. Letting the agent itself decide what to keep and what to fold feels broadly useful for any production agent fighting context bloat. MemGUI-Agent: An End-to-End Long-Horizon Mobile GUI Agent with Proactive Context Management #GUIAgents# #LLM#