Official run

Grok 4.5High

In this Twenty Questions LLM benchmark, Grok 4.5 (high) ranks 12 of 18 with a 15.17 question score and 94% success. Lower is better. The run contains 35 scored episodes across 7 subjects with 5 trials per subject. Choose a subject to inspect its attempts.

Rank 12 of 18xaiHigh

Question score15.17questions · lower is better95% CI 11.70–18.64

Success: 94%
Contract compliance: 98%
Guesser cost: $1.00
Guesser time: 40m 43s

Drill down

Subjects.

01Albert Einstein9.40 avg 02Albert Schweitzer34.20 avg 03Stephen King29.20 avg 04Garfield11.80 avg 05Achilles6.80 avg 06Genghis Khan10.40 avg 07Mario4.40 avg

Complete benchmark

Run ledger.

All Guesser and benchmark-support activity.

Total tokens: 4,985,400
Wall-clock runtime: 2h 15m
Guesser calls: 564
Completed: Jul 28, 2026, 8:29:07 PM

Reliability

Output contract breached.

10 invalid outputs affected 9 episodes and consumed 10 counted turns.

Provenance

Execution: BX-20260728-official-M0008-010
Model ID: M-0008
Benchmark: B-0001
Base seed: 0
Git commit: 37beda50c436

Recorded cost

By role.

Cost composition.Share of full-run cost

Guesser20.1%$1.00
Primary Oracle56.1%$2.80
Reviewer10.6%$0.53
Judge13%$0.65
Validator0.2%$0.01

Roles in this cost:

Guesser: Asks the questions and submits the scored guess.
Primary Oracle: Searches for evidence and proposes an answer.
Reviewer: Checks each Oracle YES or NO independently.
Judge: Decides when the Oracle and Reviewer disagree.
Validator: Checks a submitted guess against the trusted subject.

Read the role and answer-checking method →

Run configuration

Models.

Model identity, prompt versions, routing, calls, and recorded cost for the full run.

GuesserModel under test

x-ai/grok-4.5

Asks the questions and submits the scored guess.

Calls: 564
Cost: $1.0009
Reasoning: High
Routing: Exact provider · xai
Resolved model: x-ai/grok-4.5
Resolved provider: xAI
Prompt contract: stateful-category-guesser-v10-unknown-evidence-guidance
Configuration: M-0008

Resolved provider details

Legacy run. Per-call routing totals were not retained. The configured provider was xai.

Game support

Oracle and adjudication.

These models support the game. They are fixed across the run and are not under test.

Primary Oracle

openai/gpt-5.6-luna

Searches for evidence and proposes an answer.

Calls: 515
Cost: $2.8009
Reasoning: Medium
Routing: Exact provider · openai
Resolved model: openai/gpt-5.6-luna
Resolved provider: OpenAI
Prompt contract: live-web-oracle-v7-direct-negative-evidence

Resolved provider details

Legacy run. Per-call routing totals were not retained. The configured provider was openai.

Reviewer

google/gemini-3.5-flash-lite

Checks each Oracle YES or NO independently.

Calls: 484
Cost: $0.5277
Reasoning: Medium
Routing: Exact provider · google-ai-studio

Resolved provider details

Legacy run. Per-call routing totals were not retained. The configured provider was google-ai-studio.

Judge

anthropic/claude-opus-5

Decides when the Oracle and Reviewer disagree.

Calls: 54
Cost: $0.6506
Reasoning: Medium
Routing: Exact provider · anthropic

Resolved provider details

Legacy run. Per-call routing totals were not retained. The configured provider was anthropic.

Guess Validator

openai/gpt-5.6-luna

Checks a submitted guess against the trusted subject.

Calls: 37
Cost: $0.0112
Reasoning: Medium
Routing: Exact provider · openai
Resolved model: openai/gpt-5.6-luna
Resolved provider: OpenAI
Prompt contract: strict-guess-validator-v1
Configuration: gpt-5.6-luna-validator

Resolved provider details

Legacy run. Per-call routing totals were not retained. The configured provider was openai.