Head-to-head tech, judged

Two products enter.
One verdict.

Every case on The Tech Jury puts two products in the dock. Both are scored on the same named criteria — on our own tests where a product can be tested, and on published, source-labelled results where it can’t — and the jury delivers one verdict, on video, and in full right here.

The docket

Latest verdicts

The most recent cases heard by the jury. Each verdict page carries the full scorecard and the reasoning.

Case No. 006

GPT-6.2 Astra vs Sol

OpenAI's flagship costs 4x its sibling Sol — and the same as Claude Fable 5.2. The verdict: Sol for almost everyone; Fable 5.2 keeps the flagship crown.

AI ModelsNovember 19, 2026
Case No. 004

Gemini 4 Hydrogen vs Argon

Google's cheap model outscored its own flagship, 51 to 50 — and Argon's price doubles on December 1. Switch to Hydrogen for everyday work.

AI ModelsNovember 13, 2026
Case No. 003

Fable 5.2 vs Mythos 5.2

Same AI, two rulebooks. Fable is the one you can use; Mythos stays locked to vetted organizations — and should.

AI ModelsNovember 10, 2026
Case No. 002

GPT-6 Cyber: Too Dangerous to Release?

It found 948 working Linux exploits, so OpenAI rated its own model Critical and locked it to vetted defenders. A powerful defensive tool — too dangerous for general release.

AI SecurityNovember 7, 2026
Case No. 001

GPT-6.1 Terra vs the Flagships

OpenAI's $1 model matches the flagships on agent benchmarks at a sixteenth of the cost per task. Terra wins for agent work — with a human in the loop.

AI ModelsNovember 4, 2026

Procedure

How the jury works

No sponsored winners, no scores for sale. The same three steps govern every case, and the evidence is published with the verdict.

Both sides are heard

Each side is judged on the evidence: our own tests where a product can be bought and run, and published independent results — with the source named — where it can’t.

Scored on named criteria

Every case is scored on published criteria — the same yardstick for both contenders. The full scorecard appears on the verdict page and in the video.

The verdict is final

One winner, stated plainly, with the reasoning on the record. If the loser wins a category, that is on the record too. No rematches bought, no verdicts appealed by sponsors.

  • CapabilityWhat each side can actually do, on the record.
  • PerformanceBenchmark results, with the source named.
  • Price & valueTotal cost of ownership, not the teaser price.
  • AccessWho can actually get it, and on what terms.
  • Safety & privacyRisks, safeguards and fine print, read in full.
  • Staying powerWhether it will still be there next year.