# DeepSeek-R1 vs Gemini 3.8 Flash

> Gemini 3.8 Flash is the stronger model overall, scoring 59.5 to 42.0 on the Noometry Index. DeepSeek-R1 costs 1.6× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-r1-vs-gemini-3-8-flash
- Last updated: 2026-10-10
- Shared benchmarks: 24

## Summary

- They share 24 benchmarks with published results for both. DeepSeek-R1 scores higher in 0 categories and Gemini 3.8 Flash in 9 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.8 Flash leads 64.7 to 18.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 66.4% for DeepSeek-R1 and 98.9% for Gemini 3.8 Flash.
- DeepSeek-R1 is cheaper at $0.50 / $2.15 per million input/output tokens, against $0.75 / $3.75 for Gemini 3.8 Flash.
- Gemini 3.8 Flash accepts more context: 1.05M tokens versus 164K.

## Snapshot

| | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| Provider | DeepSeek | Google |
| Noometry Index | 42.0 | 59.5 |
| Rank | 110 | 13 |
| Context | 164K | 1.05M |
| Input $/M | $0.50 | $0.75 |
| Output $/M | $2.15 | $3.75 |
| Weights | Proprietary | Proprietary |

## Coding

- DeepSeek-R1: 45.9 (#63)
- Gemini 3.8 Flash: 56.4 (#17)

| Benchmark | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| SciCode | 35.7% | 56.6% |
| LMArena Coding | 1427 | 1512 |
| ALE-Bench | 804.12 | 1,270 |
| DeepSWE | — | 73.8% |
| FrontierCode | — | 41.2% |
| Aider Polyglot | 71.4% | — |
| CursorBench | — | 39.6% |
| LMArena WebDev | — | 1581 |
| FrontierSWE | — | 19.6% |
| WeirdML | 41.6% | — |
| LiveBench Coding | 66.7% | — |
| AlgoTune | 1.7 | — |

## Agentic & Tool Use

- DeepSeek-R1: 30.4 (#68)
- Gemini 3.8 Flash: 40.3 (#22)

| Benchmark | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| APEX-Agents | — | 64.3% |
| DeepResearch Bench | 35.1% | — |
| BALROG | 34.9% | — |
| GDP.pdf | — | 23.4% |
| METR Time Horizons | 53.8% | — |
| Vending-Bench 2 | — | 5,094 |

## Reasoning

- DeepSeek-R1: 18.7 (#253)
- Gemini 3.8 Flash: 64.7 (#13)

| Benchmark | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| CritPt | 1.1% | 18.3% |
| LMArena Hard Prompts | 1415 | 1506 |
| Epoch Capabilities Index | 141.3 | 157.13 |
| ARC-AGI-2 | 1.3% | — |
| SimpleBench | 40.8% | — |
| Kagi LLM Benchmark | 69.4% | — |
| NYT Connections (extended) | — | 97.4% |
| ARC-AGI-1 | 21.2% | — |
| Chess Puzzles | — | 61% |
| LiveBench Reasoning | 83.2% | — |
| Mystery Game Puzzles | — | 47% |
| DTBench | — | 95.7% |
| LiveBench Data Analysis | 69.8% | — |
| LMCA | — | 52.9% |
| Surface Evolver Bench | — | 76.9% |
| ForecastBench | 60 | — |
| LiveBench | 71.6% | — |

## Math

- DeepSeek-R1: 42.0 (#70)
- Gemini 3.8 Flash: 65.2 (#22)

| Benchmark | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.4% | 98.9% |
| LMArena Math | 1401 | 1528 |
| FrontierMath (Tiers 1-3) | — | 68.4% |
| FrontierMath Tier 4 | — | 22% |
| ProofBench | — | 48% |
| Omni-MATH | 42.4% | — |
| LiveBench Math | 80.7% | — |
| MATH Level 5 | 96.6% | — |

## Knowledge

- DeepSeek-R1: 44.3 (#81)
- Gemini 3.8 Flash: 74.9 (#2)

| Benchmark | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| GPQA Diamond | 76.3% | 95.4% |
| LMArena Expert | 1394 | 1529 |
| Humanity's Last Exam | — | 44.5% |
| SimpleQA Verified | — | 69.7% |
| MMLU-Pro | 79.3% | — |
| Confabulations | 12.7% | — |
| Vectara Hallucination Rate | 11.3% | — |
| GPQA (HELM) | 66.6% | — |

## Multimodal

- DeepSeek-R1: —
- Gemini 3.8 Flash: 40.9 (#33)

| Benchmark | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Vision | — | 1312 |
| Blueprint-Bench 2 | — | 38.6% |
| Furniture Assembly | — | 31.7% |

## Multilingual

- DeepSeek-R1: 52.6 (#80)
- Gemini 3.8 Flash: 57.9 (#5)

| Benchmark | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Non-English | 1412 | 1486 |
| LMArena Chinese | 1443 | 1547 |
| LMArena French | 1415 | 1502 |
| LMArena German | 1405 | 1470 |
| LMArena Japanese | 1394 | 1505 |
| LMArena Korean | 1359 | 1479 |
| LMArena Russian | 1423 | 1507 |
| LMArena Spanish | 1410 | 1492 |

## Instruction Following

- DeepSeek-R1: 72.2 (#138)
- Gemini 3.8 Flash: 78.2 (#11)

| Benchmark | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Instruction Following | 1382 | 1489 |
| LiveBench Instruction Following | 80.5% | — |
| IFEval | 78.4% | — |

## Long Context

- DeepSeek-R1: 45.4 (#36)
- Gemini 3.8 Flash: 46.2 (#23)

| Benchmark | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Longer Query | 1391 | 1506 |
| Fiction.LiveBench | 75% | — |

## Writing & Preference

- DeepSeek-R1: 61.5 (#84)
- Gemini 3.8 Flash: 72.1 (#16)

| Benchmark | DeepSeek-R1 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Text | 1428 | 1494 |
| LMArena Creative Writing | 1406 | 1485 |
| EQ-Bench Creative Writing | 1500 | 1748 |
| LMArena Multi-Turn | 1404 | 1503 |
| Short-Story Creative Writing | 83% | — |
| WildBench | 82.8% | — |
| LiveBench Language | 48.5% | — |

## FAQ

### Is DeepSeek-R1 better than Gemini 3.8 Flash?

Gemini 3.8 Flash is the stronger model overall, scoring 59.5 to 42.0 on the Noometry Index. DeepSeek-R1 costs 1.6× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek-R1 or Gemini 3.8 Flash?

DeepSeek-R1 is cheaper. It lists at $0.50 per million input tokens and $2.15 per million output tokens; Gemini 3.8 Flash lists at $0.75 and $3.75.

### Is DeepSeek-R1 or Gemini 3.8 Flash better for coding?

Gemini 3.8 Flash scores higher on coding benchmarks: 56.4 versus 45.9 in the Noometry coding category.

### Which has the bigger context window?

Gemini 3.8 Flash does, with 1.05M tokens against 164K.

### How many benchmarks do DeepSeek-R1 and Gemini 3.8 Flash share?

24 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and Gemini 3.8 Flash has 46.