Research2026-06-18 · source-backed
Agents' Last Exam: frontier agents pass 2.6% of the hardest economically-valuable tasks.
Story
Built with 250-plus industry experts, ALE evaluates agents on 960 expert-authored, deterministically-scored workflows across 55 industries. Agents average 26% across all tiers but 2.6% on the "last-exam" tier. The kicker: Codex with GPT-5.5 scores 82% on Terminal-Bench and 0% on Last-Exam tasks. The next time someone says agents are "job-ready," this is the number to quote back. (arXiv 2606.05405)
Related stories
Each link below shares sources, entities, or timing with this story.
Shared entities / Shared topic / What happened next / Tension
Codex on GPT-5.5 tops Terminal-Bench 2.1 at 83.4%, but the top Claude config is export-banned.
Both cover Bench, Codex, GPT, Terminal; overlapping topics (agent, codex, gpt-5); picks up the Bench thread on 2026-06-19.
Shared entities / Shared topic / Earlier coverage / Tension
GPT-5.5 Is Here. It's Fast, It's Expensive, and It Changes How You Prompt.
Both cover Bench, Codex, GPT, Terminal; overlapping topics (gpt-5, task); earlier Bench coverage from 2026-04-25.
Shared entities / Shared topic / Earlier coverage
Codex CLI on GPT-5.5 tops Terminal-Bench 2.1 at 83.4%.
Both cover Bench, Codex, GPT, Terminal; overlapping topics (codex, gpt-5); earlier Bench coverage from 2026-06-12.
GPT-5.5 powers Codex: 82.7% Terminal-Bench 2.0, native Windows, 2x Pro usage through May.
Both cover Bench, Codex, GPT, Terminal; overlapping topics (codex, gpt-5); earlier Bench coverage from 2026-05-11.
Shared entities / Shared topic / Earlier coverage / Tension
Kimi K2.6: A 1-Trillion-Parameter Open-Weights Model Just Beat GPT-5.4 on SWE-Bench Pro
Both cover Bench, GPT, Terminal; overlapping topics (agent, frontier, gpt-5); earlier Bench coverage from 2026-04-21.
GPT-5.4 Launches with Native Computer Use and 1M Context.
Both cover Built, Codex, GPT; overlapping topics (codex, gpt-5); earlier Built coverage from 2026-03-08.
Shared entities / Shared topic / Earlier coverage
Windsurf Wave 13: SWE-1.5 Free + Five Parallel Agents via Git Worktrees.
Both cover Bench, Codex, GPT; overlapping topics (agent, codex, gpt-5); earlier Bench coverage from 2026-03-23.
Shared entities / Shared topic / Earlier coverage / Tension
GPT-5.3-Codex Rollout Paused Due to Infrastructure Failure
Both cover Codex, GPT; overlapping topics (agent, back, codex, gpt-5); earlier Codex coverage from 2026-02-17.
Source trail
Entities
Provenance
- Canonical issue
- Ramsay Research Agent — June 18, 2026
- AI generated
- no
- Story unit
- 2026-06-18-agents-last-exam-frontier-agents-pass-2-6-of-the-hardest-economically-valuable-tasks
- Labels
- source-backed, canonical briefing excerpt