Gemini 3.8 Flash
| Released | 2026-09-02 |
|---|---|
| Price per 1M | $0.75 in · $3.75 out |
| Overall score | 60.7 · rank 20 |
| Context window | 1,048,576 tokens |
| Provider | |
| Weights | Proprietary |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 60.7 | 20 | 61 |
| Overall: Coding | 59.8 | 29 | 46 |
| Overall: Reasoning | 62.6 | 18 | 21 |
| Overall: Agentic work | 60.8 | 22 | 43 |
| Overall: Frontend | 57.9 | 42 | 19 |
| Coding | 59.5 | 33 | 14 |
| Agents | 60.2 | 27 | 16 |
| Reasoning | 65.9 | 9 | 13 |
| Math | 58.8 | 38 | 4 |
| Knowledge | 62.0 | 16 | 4 |
| Frontend | 56.3 | 45 | 3 |
| General | 59.3 | 34 | 5 |
| Multimodal | 62.2 | 5 | 2 |
| Value vs price | +11.7 | 21 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 59.3 ±2.3 | 15 |
| High | 60.9 ±0.6 | 46 |
| Low | 59.9 ±4.9 | 5 |
| Medium | 61.4 ±3.9 | 8 |
Evidence
Coding 59.5 ±0.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| CursorBench | 39.6% | Epoch AI | 2026-11-11 |
| FrontierSWE | 19.6% | Epoch AI | 2026-11-11 |
| DeepSWE | 73.8% | Epoch AI | 2026-11-11 |
| FrontierCode | 41.2% | Epoch AI | 2026-11-11 |
| Code Migration v1.1 | 38.6% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Coding | 65.7% | LiveBench | 2026-11-06 |
| WeirdML v3 | 7.4% | Epoch AI | 2026-11-11 |
| MirrorCode | 38.9% | Epoch AI | 2026-09-02 |
| IOI | 56.9% | Vals AI | 2026-10-29 |
| SWE-bench Multimodal | 40.8% | Vals AI | 2026-11-09 |
| SWE-bench Verified (Vals run) | 80.0% | Vals AI | 2026-09-01 |
| SciCode | 56.6% | Epoch AI | 2026-11-11 |
| LiveCodeBench (Vals run) | 89.5% | Vals AI | 2026-09-01 |
| ProgramBench | 1.0% | Vals AI | 2026-11-09 |
Agents 60.2 ±1.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Terminal-Bench 4 | 18.2% | Terminal-Bench | 2026-09-02 |
| SRE Bench | 11.1% | Vals AI | 2026-11-05 |
| Terminal-Bench 4.0 (Vals run) | 19.7% | Vals AI | 2026-11-09 |
| Terminal-Bench-Science | 8.6% | Vals AI | 2026-11-09 |
| APEX-Agents | 64.3% | Epoch AI | 2026-11-11 |
| Remote Labor Index | 5.8% | Epoch AI | 2026-11-11 |
| LiveBench 2.0 Agentic | 16.5% | LiveBench | 2026-11-06 |
| LiveBench 2.0 Browser Use | 45.1% | LiveBench | 2026-11-06 |
| Terminal-Bench 2.1 (Vals run) | 81.3% | Vals AI | 2026-10-01 |
| EMB v1.1 (Excel financial modeling) | 72.9% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Research | 79.4% | LiveBench | 2026-11-06 |
| Tax Agent Bench | 32.1% | Vals AI | 2026-11-09 |
| Finance Agent v2.1 | 65.8% | Vals AI | 2026-11-09 |
| SkillsBench | 58.0% | Vals AI | 2026-10-29 |
| Vending-Bench 2 | $5,094 | Epoch AI | 2026-11-11 |
| HLAB-Vals 1.1 (legal agent) | 12.1% | Vals AI | 2026-11-09 |
| CyberBench v1.1 | 43.8% | Vals AI | 2026-11-05 |
Reasoning 65.9 ±1.8
| Benchmark | Result | Source | Date |
|---|---|---|---|
| ARC-AGI-2 | 89.2% | Epoch AI | 2026-11-11 |
| ARC-AGI-2 (ARC Prize verified) | 89.2% | ARC Prize | 2026-11-11 |
| Extended NYT Connections | 97.4% | Lech Mazur | 2026-11-10 |
| MysteryMechanism | 36.5% | Vals AI | 2026-11-09 |
| CritPt | 18.3% | Epoch AI | 2026-11-11 |
| ARC-AGI-1 (ARC Prize verified) | 98.5% | ARC Prize | 2026-11-11 |
| ARC-AGI-1 | 98.5% | Epoch AI | 2026-11-11 |
| SimpleBench | 82.4% | SimpleBench | 2026-09-03 |
| GPQA Diamond | 95.4% | Epoch AI | 2026-09-02 |
| GPQA Diamond (Vals run) | 94.4% | Vals AI | 2026-09-01 |
| BioMysteryBench | 62.2% | Vals AI | 2026-11-09 |
| ARC-AGI-3 | 35.0% | ARC Prize | 2026-11-11 |
| Chess Puzzles | 58.0% | Epoch AI | 2026-10-13 |
| LiveBench 2.0 Reasoning | 89.3% | LiveBench | 2026-11-06 |
| Mystery Game Puzzles | 47.0% | Epoch AI | 2026-09-02 |
| LiveBench 2.0 Data Analysis | 54.0% | LiveBench | 2026-11-06 |
Math 58.8 ±1.8
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierMath Tiers 1-3 | 68.4% | Epoch AI | 2026-09-02 |
| FrontierMath Tier 4 | 22.0% | Epoch AI | 2026-09-02 |
| OTIS Mock AIME 2024-2025 | 98.9% | Epoch AI | 2026-09-02 |
| ProofBench v1.1 | 48.0% | Vals AI | 2026-11-09 |
| ProofBench | 48.0% | Epoch AI | 2026-11-11 |
Knowledge 62.0 ±2.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| SimpleQA Verified | 69.7% | Epoch AI | 2026-09-02 |
| Humanity's Last Exam | 44.5% | Epoch AI | 2026-11-11 |
| MMLU-Pro (Vals run) | 90.2% | Vals AI | 2026-09-01 |
| Legal Research Bench | 38.9% | Vals AI | 2026-11-09 |
Frontend 56.3 ±2.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LiveBench 2.0 Design | 69.7% | LiveBench | 2026-11-06 |
| Vibe Code Bench v1.1 | 78.7% | Vals AI | 2026-11-09 |
| LMArena WebDev | 1,584 | LMArena | 2026-11-09 |
General 59.3 ±1.8
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LLM Debate Benchmark | 1,505 | Lech Mazur | 2026-10-26 |
| EQ-Bench Creative Writing v3 | 1,744 | EQ-Bench | 2026-11-11 |
| LMArena Text | 1,497 | LMArena | 2026-11-09 |
| Creative Story-Writing (Lech Mazur) | 51.0% | Lech Mazur | 2026-10-23 |
| EQ-Bench Longform Creative Writing | 76.8% | EQ-Bench | 2026-11-11 |
Multimodal 62.2 ±2.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| MMMU-Pro (Vals run) | 89.1% | Vals AI | 2026-09-01 |
| LMArena Vision | 1,293 | LMArena | 2026-11-09 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| ARC-AGI-1 | 97.5% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 90.5% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 97.5% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 90.5% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 | 82.9% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 77.5% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 82.9% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 77.5% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 10.4% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 4.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 24.2% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 6.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 15.1% | ARC Prize | A stronger configuration result was used |
| CritPt | 12.3% | Epoch AI | A stronger configuration result was used |
| CritPt | 4.0% | Epoch AI | A stronger configuration result was used |
| CursorBench | 37.3% | Epoch AI | A stronger configuration result was used |
| DeepSWE | 71.0% | Epoch AI | A stronger configuration result was used |
| SciCode | 55.1% | Epoch AI | A stronger configuration result was used |
| SciCode | 55.0% | Epoch AI | A stronger configuration result was used |
Names matched to this model
gemini-3-8-flash