GPT-5.6 outscores Fable 5 on SWE-bench Verified at launch

At GPT-5.6’s public launch, its officially reported SWE-bench Verified score is higher than Fable 5’s.

44%

THE CROWD'S PULSE

YES 44%NO 56%Weighted by points on the line — 19 in.

the crowd over timen=19

44%

sealed at 47% · Jul 3now

19 on the record◎0 in playcloses Aug 14

Tech & AI

@martinssource

Exact conditions — sealed & immutable

Resolves YES if OpenAI’s officially reported SWE-bench Verified score for GPT-5.6 exceeds Anthropic’s reported Fable 5 score, measured at GPT-5.6’s launch.
Both numbers must come from the vendors’ own model cards / announcements — not third-party runs.
Voids if GPT-5.6 does not launch, or no official SWE-bench Verified number is published for it, before 2026-08-15.
Source of truth: the official model cards of GPT-5.6 and Fable 5.

sealed 2026-07-03#ffe5392bsha-256 of the sealed criteria — any edit would break it

No wagers, no odds, no cash value — accuracy only.

data: JSON · CSV · Methodology

</> Embed this call on your site

<iframe src="https://vigfi.com/embed/gpt56-beats-fable5-swebench-at-launch" width="560" height="240" style="border:0;width:100%;max-width:560px" loading="lazy" title="On the record — vigfi.com"></iframe>
<script>(function(){if(window.__vigfiEmbed)return;window.__vigfiEmbed=1;addEventListener("message",function(e){if(e.origin!=="https://vigfi.com"||!e.data||e.data.type!=="vigfi:embed-height")return;var f=document.querySelectorAll('iframe[src^="https://vigfi.com/embed/"]');for(var i=0;i<f.length;i++)if(f[i].contentWindow===e.source)f[i].style.height=e.data.height+"px"})})();</script>

GPT-5.6 outscores Fable 5 on SWE-bench Verified at launch

Exact conditions — sealed & immutable

Discussion — 0