What people are shipping with Jev.

@mvanhorn @typesafeai Holy smokes. Been using this for 3 hours wow. Context, memory, latency… pretty much all solved. Yesterday Claw was 80% of the value prop of leading Perplexit

@_pi0_ BOZO SCORE 18.6 / 100 NOT A BOZO · Has Receipts grift █░░░░░░░░░ 9 cope █░░░░░░░░░ 6 ego █░░░░░░░░░ 11 hollow █░░░░░░░░░ 8 form ███░░░░░░░ 27 "annoyingly, thi

@KinasRemek @typesafeai Ten model to początek dużych zmian. Pomijając generowanie tekstu/kodu większość tokenów agentycznych idzie na tool use i klasyfikację stanów. Jev zdaje

Hey @typesafeai ! We’ve just applied for JEV early access from Ante. Excited to try it and explore how it could fit into what we’re building around AI agents and evaluation. 👀 h

@Shasko @rafalwilinski @typesafeai Yea everyone doing something great with Jev model its great

@mikegee @typesafeai BOZO SCORE 63.9 / 100 CERTIFIED BOZO · API-Cost Accountant grift ██████░░░░ 61 cope █████░░░░░ 48 ego █████░░░░░ 45 hollow ███░░░░░░░ 31 form ███

@fatwang2ai @typesafeai @search1api_dev BOZO SCORE 24.7 / 100 NOT A BOZO · Search With Receipts grift ██░░░░░░░░ 18 cope █░░░░░░░░░ 9 ego ██░░░░░░░░ 17 hollow █░░░░░░░░░

One of the easiest @GetPeakflo use cases for Jev @typesafeai is GL / findim invoice coding on a fly. https://t.co/E4e3u9dmY0

i think ive found the best use for Jev yet fixing iphone autocorrect. no more ducking misspellings. @typesafeai https://t.co/WJpHGPrYbB

Yes yes, Jev can drive, Jev can be a programming language... but lets just skip the middle man ... Jev be the "computer" !!! Motorola 6502 emulation directly on Jev 🤣😅 @typesafea

waiting for my Jev access @typesafeai 193.6x faster, 444.6x cheaper is a bold claim i’ll take even half of it. who tried?

Acabo de despertar, soñé con Jev @typesafeai 😂😂

Progress on my ultra-fast audio concierge, connected to all my agents, now with Jev making 105 ms decisions. Playing with the model from @typesafeai has been as fun as experimentin

@aakash_env @typesafeai “LLMs generate. Jev decides” is a clean split. The 70–500 ms decision path is where this gets practical, but I’d benchmark bad-confidence fallbacks too, sin

Jev (@typesafeai) is great for profiling players! 👀 It watches you play Roulette Wars live, scores your risk-taking, and tags you as one of 4 archetypes (fortress / gambler / tac

@SUOHA_AI @jarrodwatts BOZO SCORE 99.4 / 100 终极 BOZO · 链上亏损生成器 兜售 ██████████ 99 精神胜利 ██████████ 100 自我 █████████░ 91 空洞 ██████████ 98 形式 ██████████ 99 "实盘亏了 300%，

$5 on Jev means you are set for life @typesafeai excited for trying it out! https://t.co/zfhDPWkP6z

I built a @typesafeai Jev router for Hermes. It's a plugin that you can install so that all your requests go through a jev filter first to see how strong your model needs to be. I

Jev by @typesafeai is the first model ever to truly optimize for low- and medium-entropy tasks https://t.co/W5eW7nVWyI

Jev, from TypeSafe. $42 per billion input tokens, output free. It does not write the answer. You define the question, it returns a probability. Allie Laabs (@typesafeai) walked cho