Jev AI

Jev local & open-source guide · updated October 11, 2026

Can you run Jev locally?

No. TypeSafe’s Jev is a proprietary, hosted model. Its weights are not on Hugging Face, GitHub or Ollama, so it cannot be self-hosted. What you can run locally are open-weight models that answer the same typed questions, and since Ollama 0.35 several of them sit behind Jev’s exact API: see Jev on Ollama, Kev and Jeeves. This page lists them with their licences, hardware needs and JevBench v1.7.48 standings.

Try it on this page

Run Jev 1.14 right here

Edit the example or paste your own text, then run it. Sign in to run: new accounts get 20 welcome credits, and a failed run costs nothing.

Open the full playground

Support ticket triage · Jev 1.14

1 Input

Hi, this is the third time I am writing. Our payouts have failed every day since Monday and my team cannot pay suppliers. I already re-entered the bank details twice. If this is not fixed today we will have to move to another provider.

2 Questions

  • Choice Which team should handle this ticket?
  • Yes / No Does the customer need a response today?
  • Score How frustrated is the customer?
  • Yes / No Does the customer threaten to leave or switch providers?

Is Jev open source?

The model is closed. Some of the tooling around it is public.

  • Model weights: not released. TypeSafe has not disclosed Jev’s size, base model or training data.
  • Access: a hosted API. Your text is sent to the service that runs the model.
  • SDKs: TypeSafe publishes client libraries such as @typesafe-ai/sdk. They call the hosted API; they do not contain the model.
  • “OpenJev” and similar names: third-party projects on open base models. At least ten unrelated projects use that name. See which OpenJev is which.

Facts about Jev are from TypeSafe’s public documentation as of September 18, 2026.

Open models you can run locally instead

Each answers Jev-style yes/no, choice and score questions. Standings are from JevBench v1.7.48, where hosted Jev is the open-weights board’s unranked reference at 71.4 and ranks #18 of 54 on the API board.

ModelLicenceHardware and speed notesJevBench v1.7.48
WinnowEldanRingApache-2.0, with the Gemma 4 terms that apply to derivativesA 16 GB GPU for the Q8 build; the BF16 build needs more memory or offloading#54 of 261 open weights · 72.2
CygnetblockbrainMIT shim; Gemma 4 weights under Apache-2.0 and Google’s prohibited-use policyA 48 GB GPU in the author’s runs; other cards not measured#62 of 261 open weights · 71.9
Jev-Omniakhilaaa3Apache-2.0 weights on Gemma 4 12BA CUDA GPU; about 24 GB of BF16 weights#68 of 261 open weights · 71.0
decisioAmin RoudakiApache-2.0 software; Gemma 4 weights under Google’s model termsOne NVIDIA card under vLLM; also Apple silicon, Ollama or CPU for development#70 of 261 open weights · 70.6
decider-4bMapikaApache-2.0 for weights and packageOne CUDA GPU, about 8.4 GB in bf16; also runs on CPU#129 of 261 open weights · 43.1
JevK5allebeeApache-2.0 for weights and codeYour own GPU, about 9GB in bf16; the author’s figures are on an H100 and an L40S#139 of 261 open weights · 39.2
Imajev-4BMohit GargApache-2.0 adapters and code; Apache-2.0 Qwen baseApple silicon through MLX or one CUDA GPU; 9.3 GB base weights for the 4B#153 of 261 open weights · 30.0
SemIfTheoLeeCJMIT for the project code; model weights keep upstream licencesYour own GPU; the published numbers are on an RTX 3090#183 of 261 open weights · 16.1
LayaConvai InnovationsApache-2.0, weights on Hugging Face, installable from PyPI32.8–39.5ms on a Tesla T4; 193–464ms on CPU#253 of 261 open weights · 0.0
Mercury DecideInceptionSee the comparisonSee the comparison#17 of 54 on the API board · 72.2
metask-jev-rainmetaskSee the comparisonSee the comparisonNot measured
djevMaisaCommunity djev-dev repo is Apache-2.0; Maisa’s hosted service is not yet releasedPossible: roughly one NVIDIA B200, Linux, CUDA 13, BF16Not measured
OpenJevrazorback16 / CodivApache-2.0, weights from NVIDIA and Google, about 18GB24GB NVIDIA GPU under vLLM, or Apple silicon with 16GB free under MLXOnly in thinking mode, #88 of 261

Licence and hardware details are quoted from each project’s own repository or model card, as read on November 14, 2026; follow a model’s link for the sources. The JevBench composite mixes intelligence, calibration, speed and cost, so a fast small model can rank near Jev while answering fewer decisions correctly. Read how JevBench scores are built.

Pick by the hardware you have

Start from your machine, then check the comparison for that model.

A 16 GB GPU

Winnow-12B was tested in its 8-bit build on a 16 GB card, and the same server also chats and reads images.

About 8–9 GB of GPU memory

decider-4b and JevK5 are Qwen3.5-4B models of that size. SemIf publishes its numbers on an RTX 3090.

Apple silicon

Kev-4B runs through MLX on a 32 GB Mac and Kev-0.8B on any Apple-silicon Mac; Imajev-4B runs through MLX, and razorback16’s OpenJev runs under MLX with 16 GB free. With Ollama 0.35 or later, nimble and tev1 serve Jev’s API directly (measured here), and PostHog’s Jeeves runs in FP8 on a 48 GB Mac.

A large GPU

Cygnet serves an unmodified Gemma 4 12B on vLLM and was measured on 48 GB and 96 GB cards. Jev-Omni also needs a CUDA GPU with room for a 12B model. JEMM, an Apache-2.0 Qwen3.8-27B decision model that also reads up to four screenshots, asks for a 64 GB card and has no JevBench score yet.

CPU only

Laya is a 421M encoder that runs on a CPU, but it needs fine-tuning on your labelled data before it is useful.

No hardware

Use hosted Jev. Jev AI also hosts Laya, Clef and Jev-Omni in the same playground, so you can compare before you install anything.

Hosted Jev or a local model?

The choice is mostly about data, volume and how much accuracy you need.

Reasons to run locally

  • The data cannot leave your network.
  • You already own GPU capacity, or volume is high and steady enough to keep one busy.
  • You need millisecond latency on your own hardware, offline use, or weights you can inspect and fine-tune.

Reasons to stay hosted

  • Nothing to install, operate or keep warm.
  • Jev scores 63.6 on intelligence in JevBench v1.7.48; the open 4B models score well below it.
  • A 64k-token context, where several local options stop at 16k or 32k tokens.
  • Probabilities that are calibrated as delivered. Several open models document a calibration step you must fit on your own data.

Moving existing Jev code to a local model

The request shape often carries over. The behaviour does not.

Several alternatives accept the TypeSafe-style /v1/systemone request: decider-4b and OpenJev are built to work with TypeSafe’s SDKs after a base-URL change, and JevK5 and Cygnet expose the same route. Jev AI’s own endpoint works the same way, which is how the SDK setup points an existing client here.

Before you switch, check each model’s option and context limits, rerun your labelled examples and re-fit any confidence thresholds. A threshold tuned on Jev does not transfer to another model.

Running Jev locally: FAQ

Can you run Jev locally?

No. Jev is a proprietary model that TypeSafe serves through a hosted API. TypeSafe has not published the weights, so there is nothing to download and run on your own machine. You can run an open-weight alternative locally, or call hosted Jev.

Is Jev open source?

No. The Jev model is closed. TypeSafe publishes client SDKs, but open client code does not make the remote model open source. Projects with names such as OpenJev are third-party rebuilds on other base models, not a release of Jev.

Can I download Jev from Hugging Face or GitHub?

Not the TypeSafe model. Repositories that carry the Jev name on Hugging Face or GitHub are community models, such as Jev-Omni or JevK5, built on open bases like Gemma 4 or Qwen. Check the author and the base model before assuming it is Jev.

Does Jev work with Ollama?

Not Jev itself: its weights are not public, so there is nothing to pull. Since version 0.35, released on 29 September 2026, Ollama serves Jev’s API at /v1/systemone with open decision models (nimble, tev1 and tev1:0.8b), so code written for Jev runs locally after a base-URL change. Our Jev on Ollama page measures them against hosted Jev.

What is the best open-source alternative to Jev?

It depends on your hardware and inputs. In JevBench v1.7.48 65 of 261 open-weights systems, led by Torchcast Decision 4B v1.3, score above Jev’s 71.4 reference, most of them on the strength of cheaper self-hosted serving; of the models compared here, Winnow-12B and Cygnet edge past it at #54 and #62 while trailing it on intelligence, and the 4B models are faster and lighter but trail Jev clearly on intelligence. Test the candidates on your own cases before choosing.

How large is Jev, and what hardware does it need?

TypeSafe has not disclosed Jev’s size or base model. Because it is an API, you need no GPU to use it. The open alternatives range from a 421M encoder that runs on a CPU to models of 12B and more that need a dedicated GPU.

Will my Jev code work against a local model?

Often, with a changed base URL. Several alternatives accept the same /v1/systemone request shape. Interface compatibility is not model equivalence, though: limits, calibration and answer quality differ, so re-validate your thresholds.

Set a baseline with hosted Jev first

Run your own cases in the playground, keep the answers, then compare a local model against them.

Open the playground

Every model named here belongs to its respective author and is not affiliated with Jev AI. Jev AI is a third-party platform for Jev models and is not TypeSafe’s official site. Figures can change; the linked comparison pages list their sources.