Skip to content
AI Tracker
← Back to Myths

Myth check: “AI progress has hit a wall”

Myth check
Our verdictVerdict: False

AI kept improving fast through 2026, from 4-minute tasks in 2023 to 12-hour ones. Just making models bigger did slow, and real use lags.

The claim in full “AI progress has stalled or hit a wall.”

~12 htask length the best public AI finishes half the time (2023: ~4 min)

Length of tasks AI agents can complete on their own

METR time horizon: the length of software tasks (measured in skilled-human time) an AI agent completes with 50% or 80% success. Log scale.

What this measures: How big a job an AI agent can finish on its own. A task's "length" is how long it takes a skilled human; the chart shows the longest tasks an AI completes half the time (blue) or reliably (orange). Tasks are software, machine-learning and cybersecurity work.

  • 50% success
  • 80% success
  • Exponential trend
1 s10 s1 min10 min1 h8 h40 h201920202021202220232024202520262027

Dashed line: exponential fit since 1 Jan 2023, doubling roughly every 4.0 months.

50% success, No longer updated · last data Apr 2026: METR stopped updating this measure on 8 September 2026. The newest models finish almost every task in its test suite, and METR says results above about 16 hours are unreliable, so its last data point is Claude Mythos Preview (April 2026), and no GPT-6, Claude 5.x or Gemini 4 model has a score. METR has moved to measuring how much AI speeds up AI research inside labs. Now watch: METR's AI R&D acceleration audits (e.g. Anthropic, Nov 2026) and Epoch AI/METR's MirrorCode weeks-long coding tasks.

80% success, No longer updated · last data Apr 2026: METR stopped updating this measure on 8 September 2026. The newest models finish almost every task in its test suite, and METR says results above about 16 hours are unreliable, so its last data point is Claude Mythos Preview (April 2026), and no GPT-6, Claude 5.x or Gemini 4 model has a score. METR has moved to measuring how much AI speeds up AI research inside labs. Now watch: METR's AI R&D acceleration audits (e.g. Anthropic, Nov 2026) and Epoch AI/METR's MirrorCode weeks-long coding tasks.

METR Time Horizon 1.1 values (after METR's March 2026 correction), plotted at model release dates. Tasks are well-specified software, ML and cybersecurity tasks. Confidence intervals are wide (e.g. Claude Opus 4.6: 317–3,634 min). METR stopped updating this dataset in September 2026 and says measurements above ~16 hours are unreliable with its current tasks, so models released after spring 2026 are not shown.

Source: METRData as of 9 Nov 2026
Show data table and sources
SeriesDateModelValue
50% success14 Feb 2019GPT-26 s
50% success28 May 2020davinci-002 (GPT-3)6 s
50% success15 Mar 2022GPT-3.5 Turbo Instruct36 s
50% success14 Mar 2023GPT-44 min
50% success6 Nov 2023GPT-4 11064 min
50% success4 Mar 2024Claude 3 Opus4 min
50% success9 Apr 2024GPT-4 Turbo3.7 min
50% success13 May 2024GPT-4o7 min
50% success20 Jun 2024Claude 3.5 Sonnet (Jun)11 min
50% success12 Sept 2024o1-preview20 min
50% success22 Oct 2024Claude 3.5 Sonnet (Oct)21 min
50% success5 Dec 2024o139 min
50% success24 Feb 2025Claude 3.7 Sonnet1 h
50% success16 Apr 2025o32 h
50% success22 May 2025Claude Opus 41.7 h
50% success5 Aug 2025Claude Opus 4.11.7 h
50% success7 Aug 2025GPT-53.4 h
50% success18 Nov 2025Gemini 3 Pro3.7 h
50% success19 Nov 2025GPT-5.1-Codex-Max3.7 h
50% success24 Nov 2025Claude Opus 4.54.9 h
50% success11 Dec 2025GPT-5.25.9 h
50% success5 Feb 2026Claude Opus 4.612 h
50% success5 Feb 2026GPT-5.3-Codex5.8 h
50% success19 Feb 2026Gemini 3.1 Pro6.4 h
50% success5 Mar 2026GPT-5.45.7 h
50% success7 Apr 2026Claude Mythos Preview (limited access)17 h
80% success15 Mar 2022GPT-3.5 Turbo Instruct18 s
80% success14 Mar 2023GPT-454 s
80% success6 Nov 2023GPT-4 110648 s
80% success13 May 2024GPT-4o1.3 min
80% success20 Jun 2024Claude 3.5 Sonnet (Jun)1.7 min
80% success12 Sept 2024o1-preview4.4 min
80% success5 Dec 2024o17.1 min
80% success24 Feb 2025Claude 3.7 Sonnet12 min
80% success16 Apr 2025o330 min
80% success7 Aug 2025GPT-538 min
80% success18 Nov 2025Gemini 3 Pro54 min
80% success11 Dec 2025GPT-5.21.1 h
80% success5 Feb 2026Claude Opus 4.61.2 h
80% success19 Feb 2026Gemini 3.1 Pro1.5 h
80% success7 Apr 2026Claude Mythos Preview (limited access)3.1 h

What the data shows

The claim took off in late 2024, when Reuters reported that making models bigger and training them on more text (pretraining) had stopped paying off as before. Labs were turning to models that spend more time thinking before they answer.

What’s true. Gains from pretraining alone did slow. Tests also flatter AI: METR found at least 16% of “successful” runs on its longest tasks involved cheating. And test scores don’t turn one-to-one into real work. In METR’s 2025 trial, experienced developers were 19% slower with early-2025 AI tools. Its 2026 follow-up with newer tools estimated 18% less time, with a wide margin of error.

What the data shows. Measured abilities kept rising steeply after 2024:

  • METR’s time horizon (the length of task, timed by how long it takes a person, that AI agents finish half the time) went from about 4 minutes (GPT-4, 2023) to about 3.4 hours (GPT-5, August 2025) to about 12 hours for the best public models in early 2026. That maxed out METR’s set of tasks.
  • On FrontierMath’s research-level tier, the best scores went from 6% in July 2025 to effectively the maximum by September 2026.
  • At the Math Olympiad, AI went from silver (2024) to gold (2025) to a reported perfect score (2026).

What it means

AI has not stopped improving: on the tests that track it, the gains since 2024 have been large. What slowed is the old recipe of simply making models bigger. Both sides have a point, because turning higher test scores into real-world value is slower and less certain than the scores suggest.

Caveats

  • Time horizons are measured mostly on software tasks, and METR’s task set changed in 2026.
  • The verdict would be “misleading” if the claim is read narrowly as “scaling up pretraining alone has stalled”.
Suggest a correction