Evidence ledger

TypeSafe reports Jev as materially faster than LLM workflows on its own evaluations.

The launch material reports large latency multiples; X discussion mostly repeats those figures. Trade press (The Register) relays the same vendor figures without independent measurement. One report from outside TypeSafe — a Vercel engineer, amplified by Vercel's chief executive — reports a large speedup measured on the reporter's own workload.

Vendor ClaimClaim 2

Evidence

Counterarguments

No independent benchmark in the collected dataset reproduces the headline range on representative workloads. The one report from outside the vendor covers a single workload (safety classification) compared against a single model, with no published method, so it does not generalise; the amplifying post adds reach rather than a second measurement. The status therefore stays Vendor Claim.

Open questions

What are p50/p95 latency and accuracy under equal task definitions and concurrency?

This status describes how well the claim is supported by the collected evidence, not whether it is ultimately true. A Vendor Claim is a statement TypeSafe has made about its own product; repetition by others does not upgrade it.

Sources

Limitations

No independent benchmark in the collected dataset reproduces the headline range on representative workloads. The one report from outside the vendor covers a single workload (safety classification) compared against a single model, with no published method, so it does not generalise; the amplifying post adds reach rather than a second measurement. The status therefore stays Vendor Claim.
What are p50/p95 latency and accuracy under equal task definitions and concurrency?

Where this claim matters

All claims