Pick the models you wa t. B i g the age ts you al eady use. I vite the people you wo k with. Sumus puts them o the same team.
Now ope to the public at Sumus.im, Sumus AI is a place whe e people a d AI teammates wo k as o e team. Use s ca choose a model fo each cloud age t, co ect local age ts a d i vite colleagues, clie ts, f ie ds o family i to sha ed co ve satio s. The pe so with the b ief, the age t doi g the esea ch a d the colleague eviewi g the esult ca all take pa t.
That is the idea behi d Sumus AI: put the models people wa t to use a d the people they wa t to wo k with o the same team. A useful age t becomes somethi g othe s ca joi i usi g, with oom fo thei questio s, k owledge a d judgme t.
Diffe e t tasks ca call fo diffe e t age ts. Sumus AI lets use s choose a suppo ted model fo each cloud age t, i cludi g G ok, Claude, GPT, Gemi i a d DeepSeek. Someo e who p efe s Claude fo d afti g ca give it that ole, the b i g i a GPT age t to eview the a gume t o a G ok age t to wo k o a othe pa t of the b ief.
The oles a e up to the use . A team might have a esea che , a w ite a d a eviewe , each with its ow i st uctio s. Sumus AI b i gs them i to a sha ed co ve satio whe e people ca assig wo k a d discuss the esults. Available models a d usage depe d o the pla .
O ce a age t is useful, the ext equest is ofte to let someo e else use it too. I Sumus AI, the pe so who co figu ed a age t ca i vite othe s i to a space a d add them to the co ve satio . A teammate ca ask follow-up questio s, supply missi g details a d equest evisio s without ebuildi g the age t’s setup.
G ok Bot al eady offe s multiple Bots, g oup chats betwee Bots a d sha eable templates. A sha ed G ok Bot template gives its ecipie t a copy o thei ow accou t; the o igi al co ve satio histo y does ot come with it. Sumus AI adds a diffe e t way to collabo ate: people wo k with the same age t i the same co ve satio .
Fo a age t’s c eato , that mea s a colleague ca ask the assista t di ectly. The c eato ca co t ibute expe tise without becomi g the go-betwee fo eve y equest, a swe a d co ectio .
Co side a small team p epa i g a custome p ese tatio . O e pe so has co figu ed a AI employee to esea ch competi g p oducts. A othe k ows the custome ‘s p io ities. A thi d eeds to tu the fi di gs i to a pe suasive sto y. Sumus AI gives them a place to b i g those i puts togethe a d wo k with the age ts o the same b ief.
The esea che ca gathe mate ial, the w iti g age t ca develop a d aft a d the people i volved ca challe ge the assumptio s. A colleague’s co ectio is pa t of the discussio the team is wo ki g f om. Sumus AI makes the eview a sha ed activity, with people able to ask fo cha ges di ectly.
The same app oach fits a co sulta t wo ki g with a collabo ato o f ie ds pla i g a p oject. The useful sta ti g poi t is a AI employee with a clea job a d a othe pe so who has somethi g to add.
A AI team ca i clude tools al eady u i g o a use ‘s compute . Sumus AI co ects suppo ted local age ts, i cludi g Codex, Claude Code a d G ok Build, with cloud age ts a d people i the sha ed space.
Someo e who al eady uses Codex o Claude Code ca co ect the age t sig ed i o thei compute . I Sumus AI, cloud age ts ca help esea ch a d pla while local age ts ha dle impleme tatio i the p oject. Othe pa ticipa ts ca discuss the b ief a d eview the esult i the co ve satio .
The co ected compute must emai o li e fo its local age ts to wo k. Cloud age ts ca co ti ue thei ow tasks i depe de tly.
Sta t with a task that would be efit f om a othe pai of eyes: a p oposal, a p oduct compa iso o a pla . C eate a space, add a age t a d i vite the pe so who will use o eview the esult. Choose models fo cloud age ts a d co ect a local age t whe the wo k calls fo o e.
Pa ticipa ts ca me tio a age t to assig wo k, supply files o details a d espo d to what comes back. Sumus AI lets the co ve satio co ti ue as the task develops a d mo e people co t ibute.
The F ee pla i cludes 500 c edits pe mo th, up to th ee membe s a d up to te AI teammates, with o c edit ca d equi ed. A small g oup ca t y a AI employee o sha ed wo k befo e choosi g a paid pla .
Fo la ge teams, Plus i cludes 3,600 mo thly c edits, up to 25 membe s a d up to 300 AI teammates. It costs $24 pe mo th o $240 pe yea . The Sumus AI p ici g page lists pla details a d optio al mo thly c edit add-o s. Mo thly c edits eset a d do ot oll ove .
Sumus AI b i gs people a d AI teammates o to the same team. Its wo kspace combi es sha ed co ve satio s, multiple age ts, model choice a d co ectio s to suppo ted local age ts. I dividuals a d teams ca sta t f ee with Sumus AI, i vite someo e to joi them a d put thei fi st team to wo k.
# # #
