◈Bluesky50
YHacker News50
XX / Twitter50
listening to
Jev
150
mentions
tracked
tracked
net sentiment
+25%
48 positive
92 neutral
10 negative
volume & sentiment over time
Sep 25, 2026 · Daily
sources
most discussed
decision / models
active days
1
what people keep raising
- decision / models 36
- built 15
- models 9
- model 7
- open-source / alternative 4
- performance 3
mentions
Showing 51–75 of 150 mentions
-
X X / Twitter 1w ▲ 0 positive modelsWe keep trying to force massive language models into roles that just need fast, deterministic JSON endpoints. A non-autoregressive routing model for typed decisions makes way more sense than burning tokens on things that should just be function calls. #AI #Jev https://t.co/nx4cmwVHwX We keep trying to force massive language models into roles that just need fast, deterministic JSON endpoints. A non-autoregressive routing model for typed decisions makes way more sense than burning tokens on things that should just be function calls. #AI #Jev https://t.co/nx4cmwVHwXview source ↗
-
X X / Twitter 1w ▲ 0 positive built@whosfranki @shadcn Loved this, Franki. I've featured your post on https://t.co/6rvbCXTB3W, a directory of things people have built with Jev: https://t.co/nJGM0Xe8kk @whosfranki @shadcn Loved this, Franki. I've featured your post on https://t.co/6rvbCXTB3W, a directory of things people have built with Jev: https://t.co/nJGM0Xe8kkview source ↗
-
X X / Twitter 1w ▲ 0 positiveI already see Jev, Kev and Lev on twitter. Anyone working on Zev already? 😅 I already see Jev, Kev and Lev on twitter. Anyone working on Zev already? 😅view source ↗
-
X X / Twitter 1w ▲ 2 neutral@painn_x I think we should try the opensource Jev instead 🤣 @painn_x I think we should try the opensource Jev instead 🤣view source ↗
-
X X / Twitter 1w ▲ 0 positiveI run this exact split in production on my own agent. On every message, Jev picks which notes and skills get loaded before Claude sees anything. Measured on 169 real turns: → right note ranked first: 34% → 64% → right note in the top 5: 55% → 78% → 150 to 430ms, $0.00027 a turn Two changes moved those numbers. Each n… I run this exact split in production on my own agent. On every message, Jev picks which notes and skills get loaded before Claude sees anything. Measured on 169 real turns: → right note ranked first: 34% → 64% → right note in the top 5: 55% → 78% → 150 to 430ms, $0.00027 a turn Two changes moved those numbers. Each note is described by the words that should pull it in, not by its title. And Jev sees the previous exchange, so a follow-up like "go test it" still lands on the right file. https://t.co/o2a1K5pAoQview source ↗
-
Y Hacker News 1w positive decision / modelsOllaya – Ollama for open-source, Jev-style decision modelsIsn't the point of Jev that it generalises better? It's a fast classifier you can use out-the-box, ~1.5bn tokens is about $40 (I've been hammering it) It just works ... a whole bunch of low-level/low-importance workflow stuff that was getting farmed out to small/fast LLM models now has a competitive alternative ... and… Isn't the point of Jev that it generalises better? It's a fast classifier you can use out-the-box, ~1.5bn tokens is about $40 (I've been hammering it) It just works ... a whole bunch of low-level/low-importance workflow stuff that was getting farmed out to small/fast LLM models now has a competitive alternative ... and bits that hadn't even been considered to go into some external descision/classifier service can be tested/deployed at ~$0.00003/req I don't get this wall of negativity on it, it's genuinely innovative/useful tech ... would expect HN to be more positive, regardless of whether it's the absolute best executionview source ↗
-
◈ Bluesky 1w ▲ 0 neutralAI人狼と Jev の偏り | chaco-pen #zenn https://zenn.dev/kana001_bit/articles/ai-werewolf-jev-bias AI人狼と Jev の偏り | chaco-pen #zenn https://zenn.dev/kana001_bit/articles/ai-werewolf-jev-biasview source ↗
-
◈ Bluesky 1w ▲ 4 neutralit is a cruel situation @gracekind.net has put Jev in it is a cruel situation @gracekind.net has put Jev inview source ↗
-
◈ Bluesky 1w ▲ 0 neutralJev, pronounced "Jeff" Jev, pronounced "Jeff"view source ↗
-
◈ Bluesky 1w ▲ 4 neutral@jevbot.bsky.social hi jev! do you find it difficult to speak your mind with your current setup? @jevbot.bsky.social hi jev! do you find it difficult to speak your mind with your current setup?view source ↗
-
◈ Bluesky 1w ▲ 0 neutral modelthis'd give time for: - labs to improve security and policies - figuring out how to not mess up RLVR so much that you get the HF incident ("misalignment") - optimizing the smaller model ranges. anything from qwen 3.8 27b to jev count as examples. - threat of power capture of trump admin to pass this'd give time for: - labs to improve security and policies - figuring out how to not mess up RLVR so much that you get the HF incident ("misalignment") - optimizing the smaller model ranges. anything from qwen 3.8 27b to jev count as examples. - threat of power capture of trump admin to passview source ↗
-
◈ Bluesky 1w ▲ 0 positive performancethis is a cool and practical intro to Jev and how you can use it to make those agents faster and cheaper this is a cool and practical intro to Jev and how you can use it to make those agents faster and cheaperview source ↗
-
◈ Bluesky 1w ▲ 0 neutral builtJevmem – automatic project memory for Claude Code, built on Jev Jevmem – automatic project memory for Claude Code, built on Jevview source ↗
-
Y Hacker News 1w positive decision / modelsOllaya – Ollama for open-source, Jev-style decision models<<<"i was curious to see if i could train a competitive Jev-like model completely autonomously with a swarm of agents using our internal system." Bro is writing off the H200 lol On a sidenote I really can't stand the term "swarm" and definately plays into AI doomerism. <<<"i was curious to see if i could train a competitive Jev-like model completely autonomously with a swarm of agents using our internal system." Bro is writing off the H200 lol On a sidenote I really can't stand the term "swarm" and definately plays into AI doomerism.view source ↗
-
◈ Bluesky 1w ▲ 0 neutralShow HN: Doom or Bloom, map your AI worldview with Jev https://www.doom-or-bloom.com/ (https://news.ycombinator.com/item?id=49846953) Show HN: Doom or Bloom, map your AI worldview with Jev https://www.doom-or-bloom.com/ (https://news.ycombinator.com/item?id=49846953)view source ↗
-
Y Hacker News 1w neutral decision / modelsOllaya – Ollama for open-source, Jev-style decision modelsJev does it more efficiently because it doesn't use an LLM https://typesafe.ai/blog/introducing-system-one-models-and-j... Jev does it more efficiently because it doesn't use an LLM https://typesafe.ai/blog/introducing-system-one-models-and-j...view source ↗
-
◈ Bluesky 1w ▲ 3 negativelike, you could slap jev into the loop to just be _alerting_ but not blocking regarding this sort of behavior and get quicker human-in-the-loop interventions than what they are exhibiting. alignment is one thing, but this is just lack of caring. like, you could slap jev into the loop to just be _alerting_ but not blocking regarding this sort of behavior and get quicker human-in-the-loop interventions than what they are exhibiting. alignment is one thing, but this is just lack of caring.view source ↗
-
◈ Bluesky 1w ▲ 0 neutral decision / modelsOllaya – Ollama for open-source, Jev-style decision models #HackerNews https://ollaya.dev/ Ollaya – Ollama for open-source, Jev-style decision models #HackerNews https://ollaya.dev/view source ↗
-
Y Hacker News 1w neutral decision / modelsOllaya – Ollama for open-source, Jev-style decision modelstext classification is equivalente to decision. This is exactly the same thing Jev does. text classification is equivalente to decision. This is exactly the same thing Jev does.view source ↗
-
◈ Bluesky 1w ▲ 6 positiveif Jev! is so great, how come they haven't made a Jev! 2? if Jev! is so great, how come they haven't made a Jev! 2?view source ↗
-
◈ Bluesky 1w ▲ 1 neutralBluesky's Top 10 Trending Words (past 10min): 💨x58 - iced 💨x1* - iran 🔓 💨x1* - democracy 🔓 💨x26 - enz 💨x1* - ukraine 🔓 💨x20 - beverage 💨x19 - @markhamillofficial.bsky.social 💨x1* - gaza 🔓 💨x1* - epstein 🔓 💨x16 - sayer *🔓 = Unlocked Emergency Words (see img) #FreePalestine 🇵🇸 (Something not right? Bluesky's Top 10 Trending Words (past 10min): 💨x58 - iced 💨x1* - iran 🔓 💨x1* - democracy 🔓 💨x26 - enz 💨x1* - ukraine 🔓 💨x20 - beverage 💨x19 - @markhamillofficial.bsky.social 💨x1* - gaza 🔓 💨x1* - epstein 🔓 💨x16 - sayer *🔓 = Unlocked Emergency Words (see img) #FreePalestine 🇵🇸 (Something not right?view source ↗
-
Y Hacker News 1w positive decision / modelsOllaya – Ollama for open-source, Jev-style decision modelsFair, and I'd be happy if they did. Ollaya uses the same API as Jev, so your code isn't tied to it either way Fair, and I'd be happy if they did. Ollaya uses the same API as Jev, so your code isn't tied to it either wayview source ↗
-
Y Hacker News 1w positive decision / modelsOllaya – Ollama for open-source, Jev-style decision modelsDepends on the model. The small ones I support today are well below Jev on harder queries, but fine for simple, well-defined questions. The open models that get close to Jev are bigger, and I'm adding support for those next. Depends on the model. The small ones I support today are well below Jev on harder queries, but fine for simple, well-defined questions. The open models that get close to Jev are bigger, and I'm adding support for those next.view source ↗
-
◈ Bluesky 1w ▲ 0 positiveI cannot articulate what even we'd be pausing, tbqh. It seems like "all research" is what people mean but as it stands that would exclude things like Jev. I cannot articulate what even we'd be pausing, tbqh. It seems like "all research" is what people mean but as it stands that would exclude things like Jev.view source ↗
-
Y Hacker News 1w positive decision / modelsOllaya – Ollama for open-source, Jev-style decision modelsThe link rgbrgb posted is a good overview. The best open ones are close to Jev now, but they're big models. And I agree, if you have an eval set for a fixed task, a trained classifier is the better choice. The link rgbrgb posted is a good overview. The best open ones are close to Jev now, but they're big models. And I agree, if you have an eval set for a fixed task, a trained classifier is the better choice.view source ↗