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
Pass@1; technical report says LiveCodeBench results use UI range 2025-01-01 to 2025-05-01.
条件
AI Studio API model gemini-2.5-pro; default sampling 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.
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
Pass@1; technical report says LiveCodeBench results use UI range 2025-01-01 to 2025-05-01.
条件
Stable GA API model gemini-2.5-pro; launch-day technical report 69.0%, also printed in the June 27 GA model-card column. Current 73-page report gives 74.2% for the same UI range without explaining the revision; both retained with conflict=true.
Gemini 2.5 Pro (GA) · LiveCodeBench · UI 2025-01-01 to 2025-05-01 · Scores and configurations
Gemini 2.5 Pro (GA) record for LiveCodeBench · UI 2025-01-01 to 2025-05-01. 2 score records for this release, including 2 exact values (may include historical records). Includes 3 source locations.
Gemini 2.5 Pro (GA) · LiveCodeBench scores and configurations
Metric and score
Configuration and evaluation conditions
Sources
Accuracy74.2%Reported 74.2% · Unit %
Published Unconfirmed
Conflicting results are pending review and are excluded from exact rankings.
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
Pass@1; technical report says LiveCodeBench results use UI range 2025-01-01 to 2025-05-01.
Conditions
AI Studio API model gemini-2.5-pro; default sampling settings.
Conflicting results are pending review and are excluded from exact rankings.
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
Pass@1; technical report says LiveCodeBench results use UI range 2025-01-01 to 2025-05-01.
Conditions
Stable GA API model gemini-2.5-pro; launch-day technical report 69.0%, also printed in the June 27 GA model-card column. Current 73-page report gives 74.2% for the same UI range without explaining the revision; both retained with conflict=true.