Category scores
Out of 100: 100 = the best tracked model in the category.
| Reasoning | 62 | |
| Mathematics | 42 | |
| Code | 55 | |
| Agents & real work | 70 | |
| Knowledge | 75 | |
| Long context | 92 | |
| Multimodal | 97 |
EditionUpdated daily
Ranking / Gemini 3.8 Flash
Gemini 3.8 Flash
Google · Proprietary · context 1M. Ranked #12 of 26 with 654 ERAZ points out of 1000.
ERAZ rank#12/ 26
Points654/ 1000
Coverage95%
Benchmarks won0/ 14
Input price / M$0.75
Output price / M$3.75
Points over time
| Day | Points | Change |
|---|
Results on each benchmark
| Benchmark | Score | Rank | Out of 100 | Source |
|---|---|---|---|---|
| Humanity's Last Exam | 47.8% | #13 / 26 | 78 | src ↗ |
| ARC-AGI-2 | 89.2% | #8 / 13 | 94 | src ↗ |
| ARC-AGI-3 | 10.4% | #5 / 9 | 17 | src ↗ |
| FrontierMath (Tiers 1-3) | 68.4% | #13 / 16 | 73 | src ↗ |
| FrontierMath (Tier 4) | 22% | #14 / 16 | 22 | src ↗ |
| SWE-bench Pro | not published | — | — | |
| FrontierSWE v2 | 19.6% | #12 / 14 | 30 | src ↗ |
| Terminal-Bench 2.1 | 81.3% | #5 / 12 | 93 | src ↗ |
| OSWorld 2.0 | 59% | #6 / 11 | 81 | src ↗ |
| GDPval-AA | 1545 | #20 / 23 | 63 | src ↗ |
| GPQA Diamond | 95.3% | #3 / 15 | 98 | src ↗ |
| AA-Omniscience | 29.6 | #10 / 26 | 64 | src ↗ |
| AA-LCR | 81.3% | #14 / 26 | 92 | src ↗ |
| MMMU-Pro | 85.6% | #4 / 14 | 97 | src ↗ |
Artificial Analysis Intelligence Index : 40.9 · Source page ↗
Nearby models in the ranking
</> Embed this ranking
Show Gemini 3.8 Flash's ranking on your site or in a README. The badge updates every day.
