AI Studio API model id gemini-2.5-pro; default sampling settings. Technical report states Gemini scores are pass@1 and single-attempt unless specified. Aider Polyglot averages 3 trials; SWE-bench uses the internal agentic harness; Vibe-Eval uses Gemini as judge.
未知项
Thinking budget and any non-default settings for individual evaluations are not stated except where noted in each protocol.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
MRCR-V2 8-needle; pointwise value at 1M context.
条件
Methodology uses the harder 8-needle version; pass@1 unless otherwise stated.
Gemini 2.5 Pro (GA) · MRCR-V2 · 8-needle; pointwise at 1M context · Scores and configurations
Gemini 2.5 Pro (GA) record for MRCR-V2 · 8-needle; pointwise at 1M context. 1 score record for this release, including 1 exact value (may include historical records). Includes 2 source locations.
Gemini 2.5 Pro (GA) · MRCR-V2 scores and configurations
Metric and score
Configuration and evaluation conditions
Sources
Accuracy16.4%Reported 16.4% · Unit %
Published Unconfirmed
Gemini 2.5 Pro stable API
Settings
AI Studio API model id gemini-2.5-pro; default sampling settings. Technical report states Gemini scores are pass@1 and single-attempt unless specified. Aider Polyglot averages 3 trials; SWE-bench uses the internal agentic harness; Vibe-Eval uses Gemini as judge.
Unknowns
Thinking budget and any non-default settings for individual evaluations are not stated except where noted in each protocol.
Tools
Tool access not recorded
Reasoning budget
not recorded
Framework
not recorded
Protocol
MRCR-V2 8-needle; pointwise value at 1M context.
Conditions
Methodology uses the harder 8-needle version; pass@1 unless otherwise stated.