Grok 5 METR 50% time horizon

MANIFOLD

Ṁ1.5kṀ4.9k

Jul 30

0.8%

<1.5h

1.4%

1.5h - 2h

2h - 2.5h

2.5h - 3h

3h - 3.5h

3.5h - 4h

4h - 5h

5h - 6h

6h - 7h

7h - 8h

8h - 9h

9h - 10h

10h - 11h

11h - 12h

40%

>=12h

This market will resolve to the highest 50% time horizon, as reported by METR, for any Grok 5 model released within a month of the first Grok 5 announcement.

50% time horizon is a measure of AI autonomy based on the length of tasks that AI can do: roughly, it is the time that humans take to complete tasks that an AI system can successfully do 50% of the time. See METR's Time Horizon 1.1 update for the technical definition. As of April 2026, frontier time horizons are around 12 hours, with a doubling time of roughly 4 months.

Left bounds inclusive, right bounds exclusive.

People are also trading

Grok 4.20 METR 50% time horizon

Best AI time horizon by August 2026, per METR?

Claude Opus 5 METR 50% time horizon [old version, bad buckets]

R2 / V4-Thinking METR 50% time horizon

Claude Sonnet 4.6 METR 50% time horizon

Kimi K3 Thinking METR 50% time horizon

Claude Opus 5 METR 50% time horizon

Will METR retire the 50% Time Horizon by EOY 2026

45% chance

Best METR 50% Time Horizon in 2026

Claude Mythos METR 50% time horizon

Sort by:

@Bayesian fix this market?

@jim explain

@Bayesian there's not a 50% chance that Grok 5's time-horizon is <3.5h. This market is showing the wrong probabilities.

@jim hmmm fair fair

The members of the AI futures project have given an update and they appear to now be relying on the 80% time horizon length graph from METR for their predictions rather than the 50% time horizon length graph. This implies that a 50% time horizon is not enough. While I think markets for 50% time horizons are useful, I now think that more attention needs to be paid to 80% time horizon lengths.