Just open sourced Tev1 0.8B! Weights are below. Feel free to download them. Definitely not as good as Tev1 4B or Jev, but it runs entirely locally on a mac. It can work decently well for very simple classification. huggingface.co/togethercompu…

Opus 5.5 is incredible at launch videos. I gave it an app I launched recently & it one-shotted a great launch video for me. Was very impressed with the animations & charts. I love open models and use them a lot, but I'm also a big believer in using the right model for the job

Just trained Tev1 0.8B, a tiny Jev-like classifier. Here it is running completely locally on my mac with @ollama & classifying some tasks. It's extremely fast: only ~50ms E2E latency. Video is not sped up! Releasing weights & benchmarks very soon so you can try it yourself :)

We're releasing tev1-4B-experimental, a Jev-like classifier finetuned on top of Qwen3.5 4B. We're making it available on Together serverless at $0.042/M input & $0/M output. Also releasing the data recipe & a tutorial on how to finetune your own (Tev1 cost $17 to train!).

Here's how I run my software factory: 1. During the week, I collect ideas + inspo. 2. On the weekend, I give the list to an agent to rank the best ones. 3. I spin up ~5-10 parallel agents to build POCs. 4. I kill ~60%, iterate on the better ones, and end up with 2-3 solid demos. 5. I then polish & share those demos on X. Then rinse and repeat! I still build some ideas immediately, but I'm increasingly using weekends to let agents explore ideas in parallel.

Extremely grateful! Thank you all for 100k 🙏

Jev vs GLM 5.3 at chess! Results: ◾ GLM 5.3 won by checkmate in 29 moves ◾ Jev: ~0.3s and <$0.0001 per move ◾ GLM 5.3: ~5.8s and ~$0.008 per move ◾ The whole game cost 24 cents My main takeaway is that it's often useful to use each one to their strengths: ◾ Fast, well-defined classification → specialized models like Jev ◾ Classifications that need reasoning or lookahead → LLMs like GLM 5.3 ◾ Real classification pipelines → hybrid. Jev handles the easy calls, an open model handles the hard ones. The future is multi-model!

Jev + Kimi K3 for fraud detection! TLDR: Jev classified 100 emails in 1.42 seconds, then I routed the uncertain cases to Kimi K3. The full pipeline got 96/100 correct for only ~$0.07. Video is not sped up, check out the live run! Here was my process: I gave Jev 100 emails to classify (a mix of 50 legit & 50 fraudelent emails). It classified all of them in 1.42 seconds. An underrated feature about Jev is it will give you the confidence score for a classification, so I routed any prediction under 95% confidence to Kimi K3 to be fully sure. 31 emails fell below that threshold. After routing those to Kimi K3, the combined pipeline reached 96% accuracy. The full run took 16 seconds & ~$0.07 in inference costs: - $0.068 from Kimi K3 on @togethercompute - $0.003 (1/3 of a cent) from Jev on @typesafeai. I think this is a really interesting pattern: use a fast specialized model like Jev for the narrow task, then route the uncertain cases to a larger LLM. I feel like this kind of approach could be a game changer for use cases like fraud or anything realtime. You can use the speed & low cost of Jev while having a larger LLM as a fallback to ensure high accuracy.

I used Jev to classify 1,018 AI research papers. The result: $0.08 total cost and 256ms median end-to-end latency per paper. The pipeline was: 1. Summarize each paper with DeepSeek V4 Flash 2. Send the title + summary + 24 possible topics to Jev 3. Use Jev to classify each paper 4. Visualize everything on 1kpapers.com The summaries cost $3.99 on @togethercompute. The classifications cost $0.08 on @typesafeai. So for just over $4 of inference, I ended up with a pretty useful way to explore the top AI research papers from the past year. I think this is where things are heading: different models for different parts of the workflow, instead of using one model for everything. I’m running evals on the Jev classifications before replacing the current ones, but the site is already live: 1kpapers.com

Going to be speaking at AIE New York... with my dad! He's a CPA + business professor which is perfect since the conference theme is finance in AI. Hopefully the first father and son talk at @aiDotEngineer? 😂

This model is insane at landing pages. I asked DeepSeek V4.1 Flash & Claude Fable 5 to build me a landing page for a movie theater. Fable cost $1.21 while V4.1 Flash cost 2.6 cents, making it more than 40x cheaper at similar quality. Gave both the exact same prompt!