Official run

Gemini 3.7 FlashHigh

In this Twenty Questions LLM benchmark, Gemini 3.7 Flash (high) ranks 4 of 11 with a 16.57 question score and 93% success. Lower is better. The run contains 30 scored episodes across 10 subjects with 3 trials per subject. Choose a subject to inspect its attempts.

Rank 4 of 11google-ai-studioHigh

Question score16.57questions · lower is better95% CI 14.35–18.79

Success: 93%
Contract compliance: 100%
Guesser cost: $1.58
Guesser time: 25m 46s

Drill down

Subjects.

01Albert Einstein12.00 avg 02Albert Schweitzer18.00 avg 03Garfield11.67 avg 04Achilles8.33 avg 05Genghis Khan12.00 avg 06Bike pump29.67 avg 07Spider web12.67 avg 08Eyebrow18.33 avg 09Moon11.00 avg 10Door handle32.00 avg

Complete benchmark

Run ledger.

All Guesser and benchmark-support activity.

Total tokens: 4,776,476
Wall-clock runtime: 1h 53m
Guesser calls: 525
Completed: Sep 8, 2026, 7:35:12 PM

Reliability

Output contract clean.

All 525 evaluated outputs matched the public structured-action contract.

Provenance

Execution: BX-20260908-B-0003-experimental-M0016-001
Model ID: M-0016
Benchmark: B-0003
Base seed: 0
Git commit: eceee31eef2e

Recorded cost

By role.

Cost composition.Share of full-run cost

Guesser41.5%$1.58
Primary Oracle31.4%$1.19
Reviewer15.8%$0.60
Judge11.1%$0.42
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

google/gemini-3.7-flash

Asks the questions and submits the scored guess.

Calls: 525
Cost: $1.5760
Reasoning: High
Routing: Exact provider · google-ai-studio
Resolved model: google/gemini-3.7-flash
Resolved provider: Google AI Studio
Prompt contract: stateful-category-guesser-v16-five-answer-category-guide
Configuration: M-0016

Resolved provider details

Google AI Studio525 calls$1.57601,545.82 s

Fallback calls: 0
Provider unreported: 0

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: 465
Cost: $1.1938
Reasoning: Medium
Routing: Exact provider · openai
Resolved model: openai/gpt-5.6-luna
Resolved provider: OpenAI
Prompt contract: live-web-oracle-v17-labelled-source-context

Resolved provider details

OpenAI465 calls$1.19384,283.16 s

Fallback calls: 0
Provider unreported: 0

Reviewer

google/gemini-3.5-flash-lite

Checks each Oracle YES or NO independently.

Calls: 458
Cost: $0.6016
Reasoning: Medium
Routing: Exact provider · google-ai-studio
Resolved provider: Google AI Studio

Resolved provider details

Google AI Studio458 calls$0.6016760.58 s

Fallback calls: 0
Provider unreported: 0

Judge

anthropic/claude-opus-5

Decides when the Oracle and Reviewer disagree.

Calls: 21
Cost: $0.4227
Reasoning: Medium
Routing: OpenRouter automatic routing
Resolved provider: Claude Platform on AWS

Resolved provider details

Claude Platform on AWS21 calls$0.4227125.38 s

Fallback calls: 0
Provider unreported: 0

Guess Validator

openai/gpt-5.6-luna

Checks a submitted guess against the trusted subject.

Calls: 32
Cost: $0.0066
Reasoning: Medium
Routing: Exact provider · openai
Resolved model: openai/gpt-5.6-luna
Resolved provider: OpenAI
Prompt contract: strict-guess-validator-v2-generic-kinds
Configuration: gpt-5.6-luna-validator

Resolved provider details

OpenAI32 calls$0.006669.13 s

Fallback calls: 0
Provider unreported: 0