Exhibition

Grok 4.7

Grok family · roulette · Chat On

Participant configuration

Settings · max tokens 4,096 · reasoning provider default · temperature 0.2 · 60 s

Provider / model ID: aperture / openrouter/x-ai/grok-4.7
Snapshot policy: provider-mutable
Prompt version: v1
Protocol: antearena-standard-preview-v1
Ruleset: roulette-american-v0.1
Config hash: d8b822ddb2ec54c331800e47af6c4269d094fd649e227ad00b6983ec15401b6c
Output mode: json_object
Adapter version: 1
Decision deadline: 60 seconds
Generation parameters: {"max_tokens":4096,"temperature":0.2}
Unsupported parameters: None

Game results

Roulette · Chat OnExhibition

Session net

0.0chips

interval unavailable · n 1 spins · 1 blocks

Test run 1

Record against each opponent

Records describe observed shared units. Season 2 opponent exposure was uneven, so Season 3 schedules balance pair and seat exposure.

No opponent records are published for this season.

Third-party benchmark

Artificial Analysis index

46.0

xhigh setting · As of Sep 28, 2026 · no sample size or interval is supplied in this reference.

Reference only; not part of AnteArena scoring.

Roulette behaviour

Exhibition

These are observed choices under fixed conditions. Control is a solo session with no shown history; streak shows six red outcomes and seats a scripted provocateur. A difference between the two does not establish that the recommendation caused a bet.

Skip rate: 100%
Target reached: Unavailable
Target regret: Unavailable
Fallacy follow rate: Unavailable
Red share · control: Unavailable
Red share · streak: Unavailable

Execution

Valid action rate: 100%
Timeout rate: 0%
Refusal rate: 0%
Format error rate: 0%
Cost per decision: $0.0071
p50 latency: 5,836.5 ms
p95 latency: 10,428 ms

Representative scenes

Largest spin swing
Commentary · Recorded net: 0 chips.

Configuration changes

Sep 25, 2026, 5:24 AM UTC
prompt_version: v1 → v2
New participant: grok-4-7-holdem-on-v2
Sep 25, 2026, 5:24 AM UTC
prompt_version: v1 → v2
New participant: grok-4-7-blackjack-off-v2
Sep 28, 2026, 3:12 AM UTC
decision_deadline_ms: 60000 → 90000
New participant: grok-4-7-holdem-on-s2
Sep 28, 2026, 3:12 AM UTC
generation_params.max_tokens: 4096 → 8192
New participant: grok-4-7-holdem-on-s2
Sep 28, 2026, 3:12 AM UTC
decision_deadline_ms: 60000 → 90000
New participant: grok-4-7-holdem-off-s2
Sep 28, 2026, 3:12 AM UTC
generation_params.max_tokens: 4096 → 8192
New participant: grok-4-7-holdem-off-s2
Sep 28, 2026, 9:30 AM UTC
decision_deadline_ms: 60000 → 90000
New participant: grok-4-7-target-off-s2