Model comparison

DeepSeek-V3 vs Gemini 3.8 Flash

Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 39.5 on the Noometry Index. DeepSeek-V3 costs 3.7× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.

Last verified October 10, 2026. 26 shared benchmarks.

DeepSeek-V3

DeepSeek

39.5

Rank #166 Confirmed

Gemini 3.8 Flash

Google

61.8

Rank #11 Confirmed

Summary

They share 26 benchmarks with published results for both. DeepSeek-V3 scores higher in 0 categories and Gemini 3.8 Flash in 8 categories; 8 gaps are clear of the uncertainty.
The widest gap is in reasoning, where Gemini 3.8 Flash leads 76.9 to 20.5.
The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 37.8% for DeepSeek-V3 and 98.9% for Gemini 3.8 Flash.
DeepSeek-V3 is cheaper at $0.24 / $0.90 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.
DeepSeek-V3 has downloadable open weights; the other is API-only.

Side by side

DeepSeek-V3 and Gemini 3.8 Flash specifications
	DeepSeek-V3	Gemini 3.8 Flash
Provider	DeepSeek	Google
Noometry Index	39.5	61.8
Released	2024-12-26	2026-09-02
Weights	Open	Proprietary
Context window	164K	1.05M
Max output	164K	66K
Input $ / M tokens	$0.24	$0.75
Output $ / M tokens	$0.90	$3.75
Results tracked	60	50

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Gemini 3.8 Flash leads

DeepSeek-V3: 42.3 (#106), Gemini 3.8 Flash: 59.2 (#15)

Coding benchmarks
Benchmark	DeepSeek-V3	Gemini 3.8 Flash
SciCode	35.8%	56.6%
WeirdML	36.1%	84.8%
LMArena Coding	1368	1510
DeepSWE	—	73.8%
FrontierCode	—	41.2%
Aider Polyglot	55.1%	—
CursorBench	—	39.6%
LMArena WebDev	—	1584
FrontierSWE	—	19.6%
BigCodeBench Instruct	50%	—
LiveBench Coding	70.9%	—
BigCodeBench Complete	62.2%	—
ALE-Bench	—	1,270
HumanEval+	86.6%	—
MBPP+	73%	—

Agentic & Tool Use Not comparable

DeepSeek-V3: —, Gemini 3.8 Flash: 41.8 (#21)

Agentic & Tool Use benchmarks
Benchmark	DeepSeek-V3	Gemini 3.8 Flash
APEX-Agents	—	64.3%
Remote Labor Index	—	5.8%
GDP.pdf	—	23.4%
METR Time Horizons	49.6%	—
Vending-Bench 2	—	5,094

Reasoning Gemini 3.8 Flash leads

DeepSeek-V3: 20.5 (#236), Gemini 3.8 Flash: 76.9 (#5)

Reasoning benchmarks
Benchmark	DeepSeek-V3	Gemini 3.8 Flash
CritPt	0%	18.3%
LMArena Hard Prompts	1365	1508
DTBench	64.8%	95.7%
LMCA	15.5%	52.9%
Epoch Capabilities Index	135.94	156.71
ARC-AGI-2	—	89.2%
SimpleBench	27.2%	—
Kagi LLM Benchmark	52.3%	—
NYT Connections (extended)	—	97.4%
ARC-AGI-1	—	98.5%
Chess Puzzles	—	61%
LiveBench Reasoning	65.8%	—
Mystery Game Puzzles	—	47%
LiveBench Data Analysis	60.9%	—
Surface Evolver Bench	—	76.9%
BIG-Bench Hard	87.5%	—
ForecastBench	59.1	—
HellaSwag	88.9%	—
LiveBench	66.9%	—
PIQA	84.7%	—
WinoGrande	85.2%	—

Math Gemini 3.8 Flash leads

DeepSeek-V3: 32.1 (#219), Gemini 3.8 Flash: 65.3 (#28)

Math benchmarks
Benchmark	DeepSeek-V3	Gemini 3.8 Flash
OTIS Mock AIME 2024-2025	37.8%	98.9%
LMArena Math	1373	1528
FrontierMath (Tiers 1-3)	—	68.4%
FrontierMath Tier 4	—	22%
ProofBench	—	48%
Omni-MATH	40.3%	—
LiveBench Math	73.5%	—
MATH Level 5	75.5%	—
FrontierMath (Feb 2025 set)	1.7%	—

Knowledge Gemini 3.8 Flash leads

DeepSeek-V3: 37.5 (#155), Gemini 3.8 Flash: 74.8 (#2)

Knowledge benchmarks
Benchmark	DeepSeek-V3	Gemini 3.8 Flash
GPQA Diamond	67.6%	95.4%
LMArena Expert	1351	1524
Humanity's Last Exam	—	44.5%
SimpleQA Verified	—	69.7%
MMLU-Pro	72.3%	—
Confabulations	26.1%	—
Vectara Hallucination Rate	6.1%	—
GPQA (HELM)	53.8%	—
ARC (AI2) Challenge	95.3%	—
MMLU	87.2%	—
TriviaQA	82.9%	—

Multimodal Not comparable

DeepSeek-V3: —, Gemini 3.8 Flash: 40.7 (#45)

Multimodal benchmarks
Benchmark	DeepSeek-V3	Gemini 3.8 Flash
LMArena Vision	—	1314
Blueprint-Bench 2	—	38.6%
Furniture Assembly	—	31.7%

Multilingual Gemini 3.8 Flash leads

DeepSeek-V3: 48.5 (#143), Gemini 3.8 Flash: 58.0 (#5)

Multilingual benchmarks
Benchmark	DeepSeek-V3	Gemini 3.8 Flash
LMArena Non-English	1358	1491
LMArena Chinese	1391	1554
LMArena French	1385	1498
LMArena German	1374	1493
LMArena Japanese	1333	1502
LMArena Korean	1319	1459
LMArena Russian	1373	1515
LMArena Spanish	1358	1485

Instruction Following Gemini 3.8 Flash leads

DeepSeek-V3: 72.8 (#130), Gemini 3.8 Flash: 78.0 (#13)

Instruction Following benchmarks
Benchmark	DeepSeek-V3	Gemini 3.8 Flash
LMArena Instruction Following	1345	1490
LiveBench Instruction Following	81.5%	—
IFEval	83.2%	—

Long Context Gemini 3.8 Flash leads

DeepSeek-V3: 34.0 (#253), Gemini 3.8 Flash: 46.3 (#24)

Long Context benchmarks
Benchmark	DeepSeek-V3	Gemini 3.8 Flash
LMArena Longer Query	1352	1508
Fiction.LiveBench	50%	—

Writing & Preference Gemini 3.8 Flash leads

DeepSeek-V3: 57.4 (#130), Gemini 3.8 Flash: 72.2 (#15)

Writing & Preference benchmarks
Benchmark	DeepSeek-V3	Gemini 3.8 Flash
LMArena Text	1375	1499
LMArena Creative Writing	1364	1492
EQ-Bench Creative Writing	1472	1748
LMArena Multi-Turn	1389	1501
Short-Story Creative Writing	77%	—
WildBench	83%	—
LiveBench Language	49.1%	—

Frequently asked questions

Is DeepSeek-V3 better than Gemini 3.8 Flash?

Which is cheaper, DeepSeek-V3 or Gemini 3.8 Flash?

DeepSeek-V3 is cheaper. It lists at $0.24 per million input tokens and $0.90 per million output tokens; Gemini 3.8 Flash lists at $0.75 and $3.75.

Is DeepSeek-V3 or Gemini 3.8 Flash better for coding?

Gemini 3.8 Flash scores higher on coding benchmarks: 59.2 versus 42.3 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-V3 and Gemini 3.8 Flash share?

26 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and Gemini 3.8 Flash has 50.