Is Von the same product as TypeSafe Jev?

No. Jev is TypeSafe's hosted System One service. Von is a separate product with its own weights, terms, and ops model unless explicitly documented as compatible HTTP.

Does Von beat Jev on Decision Index?

Von is not listed on the Decision Index snapshot we import; run your own eval.

Can I use the TypeSafe SDK with Von?

Von clients return fast scores but may omit metadata your Jev gates expect; test response shape in shadow mode before swapping endpoints.

Should I leave hosted Jev entirely?

Not necessarily. Teams often keep Jev for gates that need vendor support while experimenting with Von on lower-risk paths. Run shadow mode until metrics prove parity.

Why is Von outside the hub ranked table?

The ranked table lists the ten models teams most often benchmark for full System One migrations (Laya through Liquid d1). Von remains a popular runner-up when you need a fast encoder; see the hub FAQ.

madbid.lol · The paid leaderboard for AI builders. Products, agents and agencies buy their rank.

von vs Jev

Last verified: 2026-10-12.How we rank

Von publishes Apache 2.0 weights with Python and TypeScript clients. README emphasizes order-invariant option scoring and sub-25 ms inference on author hardware, with gameplay demos comparing kills and latency to hosted Jev API calls.

Von is a runner-up pick on the hub ranked table because teams often pair it with larger decoders, but Hugging Face downloads are high for its size class. Decision Index does not list Von in our snapshot, so run your own harness before you promise accuracy parity.

jev.aitools.fyi is an independent directory. We are not TypeSafe, Cloudflare, OpenAI, Perplexity, or the makers named on this page unless a listing is marked official.

Catalog listing: Von (brand-focused page; this article is the Jev comparison).

Von vs Jev at a glance

Topic	Von	TypeSafe Jev	Source
Maker	wfzyx	TypeSafe AI	GitHub: wfzyx/von
First public release	Von 1.2 weights on Hugging Face	Hosted System One (ongoing)	GitHub: wfzyx/von
License / access	Apache 2.0	Hosted API (weights not downloadable)	HF von card
Hosted price (input)	not published	$0.042 per M input tokens (vendor list, 2026-10-12)	TypeSafe models page (jev-1.13.0 price table)
API shape	Library clients; optional HTTP wrappers in community	POST /v1/systemone	Von README; Jev: TypeSafe API reference
Question types	Choice, Noul, Score style tasks	Choice, Score, Noul	Von README
Inputs	Premise text plus explicit option descriptions	Structured state string	Von README
Model sizes	Sub-billion encoder class (see card)	Hosted Jev (closed)	HF weights card
Latency	not published	not published	GitHub: wfzyx/von (no published p95 SLA cited)
Decision Index (public)	Not listed on Decision Index public board for this edition	60.11 (Decision Index 0.3.1, snapshot 2026-10-10; independent harness, not vendor-only)	Decision Index 0.3.1 public snapshot
Data handling	not published	Processed under TypeSafe terms	Self-hosted data handling not published on a single vendor page
SDKs	Python and TypeScript repositories	typesafe_sdk, gateways, MCP ecosystem	Von README
Adoption signal	883 GitHub stars (2026-10-12); 106,915 HF downloads, apache-2.0 (2026-10-12)	not published	GitHub repository (stars) · Hugging Face model card (downloads)

What Von is

Accuracy and benchmarks

Independent Decision Index edition 0.3.1 does not list Von in the snapshot we ship.

Fit and tradeoffs

Choose Von when your constraints match its deployment model: Latency below autoregressive Qwen decoders is mandatory; You route games or agents with small state and few options; You can accept DIY calibration without hosted SLAs.

Pricing and ops

Von targets sub-25 ms encoder paths on author hardware; reproduce latency and cost on your CPUs or GPUs before you promise gameplay-demo economics.

Migrating from hosted Jev

Keep your question map and threshold logic stable. Swap client configuration to point at the new base URL or SDK, then run shadow traffic until disagreement rates fall below your risk budget.

                // Before: hosted Jev
const client = new TypeSafe({ apiKey: process.env.TYPESAFE_API_KEY });
await client.systemone({ model: "jev-latest", state, questions });

// After: self-hosted System One compatible server (Kev, decider, SemIf, etc.)
const client = new TypeSafe({ baseURL: "http://localhost:8000" });
await client.systemone({ model: "von-latest", state, questions });

Ecosystem links

wfzyx/von GitHub and HF card cover clients and weights; treat hub row eleven as a speed pick, not a full migration default.

Product-specific notes

Von optimizes for order-invariant option scoring on small encoders; README gameplay demos are vendor hardware, not your SLA. Von is not listed on the public Decision Index board we import; run a private harness before accuracy promises.

Hub row 11 is an editorial speed pick, not proof Von outranks larger decoders on every adoption signal. Python and TypeScript clients exist, but you still operate the server process and secure the listen port. Teams often pair Von with a larger decoder on escalation when raw speed matters on hot paths.

Gameplay comparisons in the README are vendor-reported; reproduce on your CPUs or GPUs before finops sign-off. Order-invariant scoring helps game loops with shuffled options but still needs calibration on your real option sets.

When to choose Von

Latency below autoregressive Qwen decoders is mandatory
You route games or agents with small state and few options
You can accept DIY calibration without hosted SLAs

When to stay on Jev

You need vendor-supported System One with confidence metadata
Your tasks need large parallel question batches on one state
Decision Index scores are part of compliance evidence

FAQ

Is Von the same product as TypeSafe Jev?: No. Jev is TypeSafe's hosted System One service. Von is a separate product with its own weights, terms, and ops model unless explicitly documented as compatible HTTP.
Does Von beat Jev on Decision Index?: Von is not listed on the Decision Index snapshot we import; run your own eval.
Can I use the TypeSafe SDK with Von?: Von clients return fast scores but may omit metadata your Jev gates expect; test response shape in shadow mode before swapping endpoints.
Should I leave hosted Jev entirely?: Not necessarily. Teams often keep Jev for gates that need vendor support while experimenting with Von on lower-risk paths. Run shadow mode until metrics prove parity.
Why is Von outside the hub ranked table?: The ranked table lists the ten models teams most often benchmark for full System One migrations (Laya through Liquid d1). Von remains a popular runner-up when you need a fast encoder; see the hub FAQ.

Other Jev alternatives

Sources

Last verified 2026-10-12. Spotted an error? Contact us.

All Jev alternatives What Jev is