APEX-Agents Original Effort: High 64.3% mercor.com for Gemini 3.8 Flash (high) on APEX-Agents Original (opens in a new tab) Archived copy (opens in a new tab) Pass rate Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on APEX-Agents Original AutomationBench 1.0.6 Effort: Medium 29.7% zapier.com for Gemini 3.8 Flash (medium) on AutomationBench 1.0.6 (opens in a new tab) Archived copy (opens in a new tab) Task pass rate Reported October 7, 2026 Benchmark leaderboard Zapier Independent Zapier AutomationBench Team Record for Gemini 3.8 Flash on AutomationBench 1.0.6 AutomationBench 1.0.6 Effort: High 29.7% zapier.com for Gemini 3.8 Flash (high) on AutomationBench 1.0.6 (opens in a new tab) Archived copy (opens in a new tab) Task pass rate Reported October 7, 2026 Benchmark leaderboard Zapier Independent Zapier AutomationBench Team Record for Gemini 3.8 Flash on AutomationBench 1.0.6 BrowseComp 130-question subset — Mercor web research agent Effort: High Configuration: Mercor web research agent, 100 steps, 3 hours 83.1% mercor.com for Gemini 3.8 Flash (high) on BrowseComp 130-question subset — Mercor web research agent (opens in a new tab) Archived copy (opens in a new tab) BrowseComp Accuracy Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on BrowseComp 130-question subset — Mercor web research agent Chartography October 2026 leaderboard Effort: High 40.9% surgehq.ai for Gemini 3.8 Flash (High) on Chartography October 2026 leaderboard (opens in a new tab) Archived copy (opens in a new tab) Chartography Pass@1 Reported October 7, 2026 Benchmark leaderboard Surge AI Independent Record for Gemini 3.8 Flash on Chartography October 2026 leaderboard CharXiv Descriptive and reasoning — Mercor single-shot Effort: High Configuration: Mercor single-shot agent 95.4% mercor.com for Gemini 3.8 Flash (high) on CharXiv Descriptive and reasoning — Mercor single-shot (opens in a new tab) Archived copy (opens in a new tab) CharXiv Accuracy Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on CharXiv Descriptive and reasoning — Mercor single-shot CharXiv Reasoning — No tools Reported as: no tools Configuration: No tools 86.2% deepmind.google for Gemini 3.8 Flash (no tools) on CharXiv Reasoning — No tools (opens in a new tab) CharXiv Accuracy Reported September 2, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.8 Flash on CharXiv Reasoning — No tools CursorBench 4.0 Effort: Medium 37.3% cursor.com for Gemini 3.8 Flash (medium) on CursorBench 4.0 (opens in a new tab) Archived copy (opens in a new tab) CursorBench Accuracy Reported October 7, 2026 Benchmark leaderboard Cursor Independent Record for Gemini 3.8 Flash on CursorBench 4.0 CursorBench 4.0 Effort: High 39.6% cursor.com for Gemini 3.8 Flash (high) on CursorBench 4.0 (opens in a new tab) Archived copy (opens in a new tab) CursorBench Accuracy Reported October 7, 2026 Benchmark leaderboard Cursor Independent Record for Gemini 3.8 Flash on CursorBench 4.0 DeepSWE 1.1 Effort: High 73.7% deepmind.google for Gemini 3.8 Flash (high) on DeepSWE 1.1 (opens in a new tab) DeepSWE Pass@1 Reported September 2, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.8 Flash on DeepSWE 1.1 DeepSWE 1.1 — Mercor mini-swe-agent, 500 steps Effort: High Configuration: Mercor mini-swe-agent, 500 steps, 2 hours 71.4% mercor.com for Gemini 3.8 Flash (high) on DeepSWE 1.1 — Mercor mini-swe-agent, 500 steps (opens in a new tab) Archived copy (opens in a new tab) DeepSWE Pass@1 Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on DeepSWE 1.1 — Mercor mini-swe-agent, 500 steps Finance Agent Benchmark 2 — October 2026 evaluation Effort: High 61.4% vals.ai for Gemini 3.8 Flash (high) on Finance Agent Benchmark 2 — October 2026 evaluation (opens in a new tab) Archived copy (opens in a new tab) Finance Agent Accuracy Reported October 6, 2026 Benchmark leaderboard Vals AI Independent Record for Gemini 3.8 Flash on Finance Agent Benchmark 2 — October 2026 evaluation FrontierSWE 2 — October 2026 evaluation Configuration: Proximus harness, 5 trials, 20-hour budget 19.6% frontierswe.com for Gemini 3.8 Flash on FrontierSWE 2 — October 2026 evaluation (opens in a new tab) Archived copy (opens in a new tab) FrontierSWE Score Reported October 7, 2026 Benchmark leaderboard Proximal Independent Record for Gemini 3.8 Flash on FrontierSWE 2 — October 2026 evaluation GDP.pdf 35.0% deepmind.google for Gemini 3.8 Flash on GDP.pdf GDP.pdf (opens in a new tab) GDP.pdf All Pass Rate Reported September 2, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.8 Flash on GDP.pdf GDP.pdf GPQA Diamond — Epoch AI harness Effort: High Configuration: Epoch AI harness 95.4% epoch.ai for Gemini 3.8 Flash (high) on Graduate-Level Google-Proof Q&A Diamond — Epoch AI harness (opens in a new tab) Archived copy (opens in a new tab) GPQA Diamond Accuracy Reported October 7, 2026 Evaluated September 2, 2026 at 5:15 PM UTC Benchmark leaderboard Epoch AI Independent Record for Gemini 3.8 Flash on Graduate-Level Google-Proof Q&A Diamond — Epoch AI harness GPQA Diamond — Mercor single-shot Effort: High Configuration: Mercor single-shot agent 93.9% mercor.com for Gemini 3.8 Flash (high) on Graduate-Level Google-Proof Q&A Diamond — Mercor single-shot (opens in a new tab) Archived copy (opens in a new tab) GPQA Diamond Accuracy Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on Graduate-Level Google-Proof Q&A Diamond — Mercor single-shot GPQA Diamond — Vals AI harness Effort: High Configuration: Vals AI harness: mean of zero-shot and few-shot CoT 94.4% vals.ai for Gemini 3.8 Flash (high) on Graduate-Level Google-Proof Q&A Diamond — Vals AI harness (opens in a new tab) Archived copy (opens in a new tab) GPQA Diamond Accuracy Reported September 1, 2026 Benchmark leaderboard Vals AI Independent Record for Gemini 3.8 Flash on Graduate-Level Google-Proof Q&A Diamond — Vals AI harness HLAB October 2026 evaluation Effort: High 10.0% vals.ai for Gemini 3.8 Flash (high) on Harvey's Legal Agent Benchmark October 2026 evaluation (opens in a new tab) Archived copy (opens in a new tab) HLAB Task Pass Rate Reported October 6, 2026 Benchmark leaderboard Vals AI Independent Record for Gemini 3.8 Flash on Harvey's Legal Agent Benchmark October 2026 evaluation HLE-Verified Full 1,811-item verified set 54.9% deepmind.google for Gemini 3.8 Flash on HLE-Verified Full 1,811-item verified set (opens in a new tab) HLE-Verified Accuracy Reported September 2, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.8 Flash on HLE-Verified Full 1,811-item verified set Humanity's Last Exam Full set — Scale AI evaluation 44.5% labs.scale.com for Gemini 3.8 Flash on Humanity's Last Exam Full set — Scale AI evaluation (opens in a new tab) Archived copy (opens in a new tab) Humanity's Last Exam Accuracy Reported October 3, 2026 Benchmark leaderboard Scale AI Independent Record for Gemini 3.8 Flash on Humanity's Last Exam Full set — Scale AI evaluation Humanity's Last Exam Text only — Mercor single-shot Effort: High Configuration: Mercor single-shot agent 47.0% mercor.com for Gemini 3.8 Flash (high) on Humanity's Last Exam Text only — Mercor single-shot (opens in a new tab) Archived copy (opens in a new tab) Humanity's Last Exam Accuracy Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on Humanity's Last Exam Text only — Mercor single-shot Humanity's Last Exam Text only — Scale AI evaluation 45.0% labs.scale.com for Gemini 3.8 Flash on Humanity's Last Exam Text only — Scale AI evaluation (opens in a new tab) Archived copy (opens in a new tab) Humanity's Last Exam Accuracy Reported October 7, 2026 Benchmark leaderboard Scale AI Independent Record for Gemini 3.8 Flash on Humanity's Last Exam Text only — Scale AI evaluation LVBench 1024 frames — Agentic 87.8% deepmind.google for Gemini 3.8 Flash on LVBench 1024 frames — Agentic (opens in a new tab) LVBench Accuracy Reported September 2, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.8 Flash on LVBench 1024 frames — Agentic LVBench 1024 frames — Static 87.1% deepmind.google for Gemini 3.8 Flash on LVBench 1024 frames — Static (opens in a new tab) LVBench Accuracy Reported September 2, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.8 Flash on LVBench 1024 frames — Static MedXpertQA MM — Mercor single-shot Effort: High Configuration: Mercor single-shot agent 84.7% mercor.com for Gemini 3.8 Flash (high) on MedXpertQA MM — Mercor single-shot (opens in a new tab) Archived copy (opens in a new tab) Accuracy Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on MedXpertQA MM — Mercor single-shot MMLU-Pro Original — Mercor single-shot Effort: High Configuration: Mercor single-shot agent 90.7% mercor.com for Gemini 3.8 Flash (high) on MMLU-Pro Original — Mercor single-shot (opens in a new tab) Archived copy (opens in a new tab) Accuracy Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on MMLU-Pro Original — Mercor single-shot MMLU-Pro Original — Vals AI harness Effort: High Configuration: Vals AI harness 90.2% vals.ai for Gemini 3.8 Flash (high) on MMLU-Pro Original — Vals AI harness (opens in a new tab) Archived copy (opens in a new tab) Accuracy Reported September 1, 2026 Benchmark leaderboard Vals AI Independent Record for Gemini 3.8 Flash on MMLU-Pro Original — Vals AI harness MMMU-Pro 1,730 questions — Mercor single-shot Effort: High Configuration: Mercor single-shot agent 85.3% mercor.com for Gemini 3.8 Flash (high) on MMMU-Pro 1,730 questions — Mercor single-shot (opens in a new tab) Archived copy (opens in a new tab) MMMU-Pro accuracy Setting: Not stated Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on MMMU-Pro 1,730 questions — Mercor single-shot MMMU-Pro Vals AI harness Effort: High Configuration: Vals AI harness 89.1% vals.ai for Gemini 3.8 Flash (high) on MMMU-Pro Vals AI harness (opens in a new tab) Archived copy (opens in a new tab) MMMU-Pro accuracy Setting: Standard (4 options) Reported September 1, 2026 Benchmark leaderboard Vals AI Independent Record for Gemini 3.8 Flash on MMMU-Pro Vals AI harness Riemann-bench October 2026 leaderboard Effort: High 51.2% surgehq.ai for Gemini 3.8 Flash (High) on Riemann-bench October 2026 leaderboard (opens in a new tab) Archived copy (opens in a new tab) Riemann-bench Pass@1 Reported October 7, 2026 Benchmark leaderboard Surge AI Independent Record for Gemini 3.8 Flash on Riemann-bench October 2026 leaderboard SimpleQA Verified Original — Epoch AI harness Effort: High Configuration: Epoch AI harness 69.7% epoch.ai for Gemini 3.8 Flash (high) on SimpleQA Verified Original — Epoch AI harness (opens in a new tab) Archived copy (opens in a new tab) Accuracy Reported October 7, 2026 Evaluated September 2, 2026 at 5:15 PM UTC Benchmark leaderboard Epoch AI Independent Record for Gemini 3.8 Flash on SimpleQA Verified Original — Epoch AI harness SWE Atlas Codebase QnA — Public; mini-swe-agent Configuration: mini-swe-agent harness 47.0% labs.scale.com for Gemini 3.8 Flash on SWE Atlas Codebase QnA — Public; mini-swe-agent (opens in a new tab) Archived copy (opens in a new tab) Mean pass@1 Reported October 7, 2026 Benchmark leaderboard Scale AI Independent Record for Gemini 3.8 Flash on SWE Atlas Codebase QnA — Public; mini-swe-agent SWE Atlas Refactoring — mini-swe-agent Configuration: mini-swe-agent harness 44.8% labs.scale.com for Gemini 3.8 Flash on SWE Atlas Refactoring — mini-swe-agent (opens in a new tab) Archived copy (opens in a new tab) Mean pass@1 Reported October 7, 2026 Benchmark leaderboard Scale AI Independent Record for Gemini 3.8 Flash on SWE Atlas Refactoring — mini-swe-agent SWE-bench Multilingual — Mercor mini-swe-agent, 1000 steps Effort: High Configuration: Mercor mini-swe-agent, 1000 steps, 3 hours 85.8% mercor.com for Gemini 3.8 Flash (high) on SWE-bench Multilingual — Mercor mini-swe-agent, 1000 steps (opens in a new tab) Archived copy (opens in a new tab) Resolved Tasks Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on SWE-bench Multilingual — Mercor mini-swe-agent, 1000 steps SWE-bench Verified — Mercor mini-swe-agent, 1000 steps Effort: High Configuration: Mercor mini-swe-agent, 1000 steps, 3 hours 80.6% mercor.com for Gemini 3.8 Flash (high) on SWE-bench Verified — Mercor mini-swe-agent, 1000 steps (opens in a new tab) Archived copy (opens in a new tab) Resolved Tasks Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on SWE-bench Verified — Mercor mini-swe-agent, 1000 steps SWE-bench Verified — Vals AI mini-swe-agent Effort: High Configuration: mini-swe-agent harness 80.0% vals.ai for Gemini 3.8 Flash (high) on SWE-bench Verified — Vals AI mini-swe-agent (opens in a new tab) Archived copy (opens in a new tab) Resolved Tasks Reported September 1, 2026 Benchmark leaderboard Vals AI Independent Record for Gemini 3.8 Flash on SWE-bench Verified — Vals AI mini-swe-agent Terminal-Bench 2.1 89.4% deepmind.google for Gemini 3.8 Flash on Terminal-Bench 2.1 (opens in a new tab) Terminal-Bench Accuracy Reported September 2, 2026 Model card Google DeepMind Self-reported Record for Gemini 3.8 Flash on Terminal-Bench 2.1 Terminal-Bench 2.1 — Vals AI Terminus 2 Effort: High Configuration: Terminus-2 harness 81.3% vals.ai for Gemini 3.8 Flash (high) on Terminal-Bench 2.1 — Vals AI Terminus 2 (opens in a new tab) Archived copy (opens in a new tab) Terminal-Bench Accuracy Reported September 28, 2026 Benchmark leaderboard Vals AI Independent Record for Gemini 3.8 Flash on Terminal-Bench 2.1 — Vals AI Terminus 2 Terminal-Bench 4.0 19.1% deepmind.google for Gemini 3.8 Flash on Terminal-Bench 4.0 (opens in a new tab) Terminal-Bench Accuracy Reported September 2, 2026 Model card Google DeepMind Self-reported Terminal-Bench Team Record for Gemini 3.8 Flash on Terminal-Bench 4.0 Terminal-Bench 4.0 — Mercor mini-swe-agent Effort: High Configuration: Mercor mini-swe-agent, 1000 steps, 8 hours 12.1% mercor.com for Gemini 3.8 Flash (high) on Terminal-Bench 4.0 — Mercor mini-swe-agent (opens in a new tab) Archived copy (opens in a new tab) Terminal-Bench Accuracy Reported October 7, 2026 Benchmark leaderboard Mercor Independent Record for Gemini 3.8 Flash on Terminal-Bench 4.0 — Mercor mini-swe-agent Terminal-Bench 4.0 — mini-swe-agent Effort: High Configuration: mini-swe-agent harness 19.1% tbench.ai for Gemini 3.8 Flash (high) on Terminal-Bench 4.0 — mini-swe-agent (opens in a new tab) Archived copy (opens in a new tab) Terminal-Bench Accuracy Reported September 2, 2026 Benchmark leaderboard Terminal-Bench Independent Terminal-Bench Team Record for Gemini 3.8 Flash on Terminal-Bench 4.0 — mini-swe-agent Terminal-Bench 4.0 — Vals AI mini-swe-agent Effort: High Configuration: mini-swe-agent harness 19.2% vals.ai for Gemini 3.8 Flash (high) on Terminal-Bench 4.0 — Vals AI mini-swe-agent (opens in a new tab) Archived copy (opens in a new tab) Terminal-Bench Accuracy Reported October 6, 2026 Benchmark leaderboard Vals AI Independent Record for Gemini 3.8 Flash on Terminal-Bench 4.0 — Vals AI mini-swe-agent Vals Index 2.1 Effort: High 54.8% vals.ai for Gemini 3.8 Flash (high) on Vals Index 2.1 (opens in a new tab) Archived copy (opens in a new tab) Vals Index GDP-weighted Accuracy Reported October 6, 2026 Benchmark leaderboard Vals AI Independent Record for Gemini 3.8 Flash on Vals Index 2.1 Vibe Code Bench 1.1 — OpenHands, October 2026 evaluation Effort: High Configuration: OpenHands harness 78.7% vals.ai for Gemini 3.8 Flash (high) on Vibe Code Bench 1.1 — OpenHands, October 2026 evaluation (opens in a new tab) Archived copy (opens in a new tab) Vibe Code Bench Score Reported October 6, 2026 Benchmark leaderboard Vals AI Independent Record for Gemini 3.8 Flash on Vibe Code Bench 1.1 — OpenHands, October 2026 evaluation