Bounded prime gap constant

Description of constant

Let $p_n$ denote the $n$-th prime. The bounded prime gap constant is

\[C_{88a} \;=\; H_1 \;:=\; \liminf_{n \to \infty} (p_{n+1} - p_n),\]

the least limit point of the sequence of gaps between consecutive primes. That $H_1$ is finite — that some bounded gap recurs infinitely often — was proved by Zhang [Zha14] in 2013; the twin prime conjecture is exactly the assertion $H_1 = 2$.

More generally one writes $H_m := \liminf_{n\to\infty}(p_{n+m} - p_n)$ for $m \ge 1$. Maynard [May15] proved $H_m$ finite for every $m$; the growth rate of $H_m$ in $m$ is recorded separately as $C_{88b}$.

Known upper bounds

Bound Reference Comments
$70\,000\,000$ [Zha14] The first finiteness proof, via a bounded-level equidistribution estimate for smooth moduli beyond the Bombieri–Vinogradov range.
$4680$ [Pol14a] Polymath8a, by optimizing Zhang’s equidistribution estimates.
$600$ [May15] Maynard, via a multidimensional generalization of the Selberg sieve, using only Bombieri–Vinogradov. Obtained independently by Tao.
$246$ [Pol14b] Polymath8b, generalizing the Maynard sieve further together with extensive numerical work. Stood as the record for twelve years.
$240$ [Sta26] Stadlmann: “the Bombieri–Vinogradov theorem can be combined with newer equidistribution estimates for smooth moduli”, the latter being the exponent of distribution $\tfrac12 + \tfrac1{40}$ of [Sta25]. Preprint of 31 August 2026.
$186$ [OAI26] Establishes $\mathrm{DHL}[40,2]$: every admissible $40$-tuple has infinitely many translates containing at least two primes. Combines the equidistribution estimates of [Pol14a] and [Sta25] with factorization conditions making suitable least common multiples of divisor products triply densely divisible, which widens the support of the multidimensional Selberg sieve, together with an improved numerical optimization. Applied to the admissible $40$-tuple $\{0, 2, 6, \dots, 182, 186\}$ of diameter $186$. Preprint of 30 August 2026, one day before [Sta26], which it describes as independent concurrent work; the proof is attributed to the model GPT 6 Astra. See the note on its formalization below.

Known lower bounds

Bound Reference Comments
$2$ Trivial $p_{n+1} - p_n \ge 2$ for every $n \ge 2$. Equality, $H_1 = 2$, is the twin prime conjecture.

References

Contribution notes

Prepared with assistance from Claude Opus 5, which read the arXiv abstract and HTML introduction of [Sta26], the abstracts of [Sta25] and [Pol14b], and confirmed the bibliographic details of every reference through Crossref. The conditional bounds and the $70\,000\,000 \to 4680 \to 600 \to 246$ chain are quoted from the abstract of [Pol14b]; the $240$ and its provenance from [Sta26]. No bound was independently verified.

The $186$ row was added in #166 by @andrew-yoo and filled out with assistance from Claude Opus 5, which read [OAI26] itself and the README.md and formalization.yaml of the accompanying Lean repository. The axiom list, the zero-sorry claim and the self-assessed review status are as reported there; neither the Lean build nor the numerical certificate was re-run here.