Start free trial

Corrected 8 Nov 2026: read the correction

TRUE, BUTMoved Ruler

Per token, yes.

Per finished task, one index puts it far above a rival at the same price.

“Claude Haiku 5.5 is the cheapest, fastest and most capable small model Anthropic has released, and costs around 75% less to run than Haiku 4.5.”

Anthropic, launch announcement, top of Hacker News (Oct 7, 2026)

11 receipts5 support3 refute3 context

Issue 388 October 20263 min
A fuel pump marked with a dollar sign stands in front of a small silver car, next to a fuel gauge that is nearly empty.
AI illustration

Artificial Analysis confirms the per-token price matches GPT-6 Luna and rises 5x above 100k tokens.

The twist

Haiku 5.5 burns far more tokens, so on the Vals Index it costs $3.00 per test against $0.43 for Luna, for 54.44% against 51.22%.

75%Anthropic's average saving to run Haiku 5.5 against Haiku 4.5
$3.00Claude Haiku 5.5 cost per test on the Vals Index
$0.43GPT-6 Luna cost per test on the same index
54.44%Claude Haiku 5.5 score on the Vals Index

There’s more to this story.

Membership opens the full investigation, the strongest counterargument and what to do with what you’ve learned.

30 days free, then A$89 a year. Payment card required. Cancel before the trial ends to avoid the first charge.

Start free trial →
  1. At sign-up: 30 days free. Payment card required. One introductory trial per customer.
  2. Your trial ends 30 days after you start. Unless you cancel before then in Account → Manage subscription, we charge A$89 for the first year. It renews automatically at A$89/yr until cancelled.
  3. You can ask for a full refund within 14 days after any annual payment, renewals included, with no reason needed, by emailing hello@bskiller.com. This voluntary refund does not limit your rights under the Australian Consumer Law.

By starting your trial, you agree to the Terms.

Inferno Tech Pty Ltd t/a BS Killer, ABN 27 647 413 474

Already a member? Sign in

The trick has a name

We call it Moved Ruler: two methods, two answers, one of them quoted. You'll see it again. Learn to spot it →

Say this in tomorrow's meeting“Haiku 5.5 is much cheaper than Haiku 4.5, but it matches GPT-6 Luna only per token; on the Vals Index it costs $3.00 per test against $0.43.”

Receipts

  1. Supports anthropic.com: the cheapest, fastest, and most capable small model
  2. Supports anthropic.com: Haiku 5.5 is available at a much lower price than Haiku 4.5.
  3. Supports anthropic.com: On average, it now costs around 75% less to run.
  4. Context anthropic.com: 50% lower for requests over 100,000 tokens
  5. Context anthropic.com: which means it uses slightly more tokens per task
  6. Supports artificialanalysis.ai: the same as GPT-6 Luna and 10% of the previous Haiku model
  7. Context artificialanalysis.ai: However, this pricing rises 5x to $0.50/$2.50 above 100k.
  8. Refutes artificialanalysis.ai: uses ~162k output tokens per Intelligence Index task, ~3x GPT-6 Luna (max, ~50k)
  9. Supports artificialanalysis.ai: At max effort Haiku 5.5 sits slightly ahead of models such as GLM-5.3 Flash (42), Gemini 3.8 Flash (41) and GPT-6 Luna (38)
  10. Refutes web.archive.org: 18 Claude Haiku 5.5 54.44% $3.00 $0.1 / $0.5
  11. Refutes web.archive.org: 28 GPT-6 Luna 51.22% $0.43 $0.1 / $0.5

Open the Receipts Pack → What each source proves, every figure traced, and what would change our verdict.

Checked by AI at the BS Killer desk · ruled 8 Oct 2026 · How we check →

Get tomorrow's check in your inbox.

One AI claim a night, checked against independent sources, receipts attached. Free.

Free email updates. Unsubscribe any time.

This story is a stable, citable object. If you can falsify a verdict, tell us. Corrections are loud here.