Gemini 3.8 Flash
Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows
Shogo Agent Index
Supported#21 of 9256.2range 48.7–63.7 · 34% of weight measured
Agentic coding
63
43% measured
Long-horizon autonomy
64
30% measured
Knowledge-work agents
47
52% measured
Computer use
44
estimated
Security & context learning
53
estimated
Agentic benchmarks
| Benchmark | Score | Normalised | Index weight | Harness / source |
|---|---|---|---|---|
| SWE-bench Verified | est. | 8% | predicted 82 | |
| Terminal-Bench 2.0 | est. | 8% | predicted 83 | |
| DeepSWE | 73.8% | 6% | mini-swe-agent · $2.36/task | |
| FrontierSWE | 19.6% | 5% | proximus · $38.75/task | |
| CursorBench | 39.6% | 4% | source · $4.70/task | |
| GSO-Bench | est. | 4% | predicted 43 | |
| METR Time Horizons | est. | 10% | predicted 79 | |
| Vending-Bench 2 | $5,094 | 6% | source | |
| PostTrainBench | est. | 4% | predicted 29 | |
| APEX-Agents | 64.3% | 8% | source | |
| GDPval | est. | 7% | predicted 55 | |
| Remote Labor Index | 5.8% | 5% | source | |
| DeepResearch Bench | est. | 5% | predicted 50 | |
| OSWorld 2.0 | est. | 6% | predicted 16 | |
| OSWorld | est. | 4% | predicted 85 | |
| Cybench | est. | 5% | predicted 83 | |
| CL-bench | est. | 5% | predicted 24 |
Other benchmarks
| Benchmark | Category | Score | Source |
|---|---|---|---|
| ARC-AGI | Reasoning | 98.5% | source |
| ARC-AGI-2 | Reasoning | 89.2% | source |
| Chess Puzzles | Reasoning | 58.0% | source |
| CritPt | Reasoning | 18.3% | source |
| DTBench | Knowledge | 95.7% | source |
| FrontierCode | Coding | 41.2% | devin-cli |
| FrontierMath Tiers 1-3 | Math | 68.4% | source |
| FrontierMath Tier 4 | Math | 22.0% | source |
| GeoBench | Multimodal | 86.0% | source |
| GPQA Diamond | Reasoning | 95.4% | source |
| Humanity's Last Exam | Reasoning | 44.5% | source |
| LMCA | Knowledge | 52.9% | source |
| MirrorCode | Coding | 38.9% | source |
| OTIS Mock AIME | Math | 98.9% | source |
| ProofBench | Math | 48.0% | source |
| SciCode | Coding | 56.6% | source |
| SimpleBench | Reasoning | 82.4% | source |
| SimpleQA Verified | Knowledge | 69.7% | source |
| WebDev Arena | Coding | 1568 | source |
| WeirdML v3 | Coding | 7.4% | gemini_cli |
API hosts
Live prices and uptime for every host serving Gemini 3.8 Flash via OpenRouter.
| Host | Input | Output | Cache read | Agent $/M | Uptime 24h | Quant |
|---|---|---|---|---|---|---|
| $0.38 | $1.88 | $0.037 | $0.17 | 72.0% | — | |
| Google AI Studio | $0.38 | $1.88 | $0.037 | $0.17 | 100.0% | — |
| $0.75 | $3.75 | $0.075 | $0.35 | 100.0% | — | |
| Google AI Studio | $0.75 | $3.75 | $0.075 | $0.35 | 100.0% | — |
| $1.35 | $6.75 | $0.14 | $0.62 | 100.0% | — | |
| Google AI Studio | $1.35 | $6.75 | $0.14 | $0.62 | 100.0% | — |