Model comparison

DeepSeek-V3.2-Exp vs Gemini 3.8 Flash

Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 44.3 on the Noometry Index. DeepSeek-V3.2-Exp costs 5.2× 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. 34 shared benchmarks.

DeepSeek-V3.2-Exp

DeepSeek

44.3

Rank #78 Confirmed

Gemini 3.8 Flash

Google

61.8

Rank #11 Confirmed

Summary

They share 34 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 1 category and Gemini 3.8 Flash in 8 categories; 9 gaps are clear of the uncertainty.
The widest gap is in reasoning, where Gemini 3.8 Flash leads 76.9 to 22.1.
The biggest single-benchmark swing is ARC-AGI-2: 4% for DeepSeek-V3.2-Exp and 89.2% for Gemini 3.8 Flash.
DeepSeek-V3.2-Exp is cheaper at $0.26 / $0.38 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-V3.2-Exp has downloadable open weights; the other is API-only.

Side by side

DeepSeek-V3.2-Exp and Gemini 3.8 Flash specifications
	DeepSeek-V3.2-Exp	Gemini 3.8 Flash
Provider	DeepSeek	Google
Noometry Index	44.3	61.8
Released	2025-09-29	2026-09-02
Weights	Open	Proprietary
Context window	164K	1.05M
Max output	66K	66K
Input $ / M tokens	$0.26	$0.75
Output $ / M tokens	$0.38	$3.75
Results tracked	49	50

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

Category by category

Coding Gemini 3.8 Flash leads

DeepSeek-V3.2-Exp: 46.5 (#65), Gemini 3.8 Flash: 59.2 (#15)

Coding benchmarks
Benchmark	DeepSeek-V3.2-Exp	Gemini 3.8 Flash
LMArena WebDev	1362	1584
SciCode	38.9%	56.6%
WeirdML	39.5%	84.8%
LMArena Coding	1454	1510
DeepSWE	—	73.8%
FrontierCode	—	41.2%
SWE-bench Verified (bash only)	70%	—
Aider Polyglot	74.2%	—
CursorBench	—	39.6%
SWE-bench Multilingual	59%	—
FrontierSWE	—	19.6%
ALE-Bench	—	1,270

Agentic & Tool Use Gemini 3.8 Flash leads

DeepSeek-V3.2-Exp: 32.7 (#59), Gemini 3.8 Flash: 41.8 (#21)

Agentic & Tool Use benchmarks
Benchmark	DeepSeek-V3.2-Exp	Gemini 3.8 Flash
APEX-Agents	21.3%	64.3%
Vending-Bench 2	1,034	5,094
Terminal-Bench	39.6%	—
Berkeley Function Calling Leaderboard	56.7%	—
Remote Labor Index	—	5.8%
TheAgentCompany	42.9%	—
GDP.pdf	—	23.4%

Reasoning Gemini 3.8 Flash leads

DeepSeek-V3.2-Exp: 22.1 (#208), Gemini 3.8 Flash: 76.9 (#5)

Reasoning benchmarks
Benchmark	DeepSeek-V3.2-Exp	Gemini 3.8 Flash
ARC-AGI-2	4%	89.2%
NYT Connections (extended)	36.7%	97.4%
ARC-AGI-1	57%	98.5%
CritPt	2.9%	18.3%
Chess Puzzles	14%	61%
LMArena Hard Prompts	1434	1508
DTBench	87.7%	95.7%
LMCA	29.1%	52.9%
Epoch Capabilities Index	146.27	156.71
Kagi LLM Benchmark	52.2%	—
Thematic Generalization	65%	—
Mystery Game Puzzles	—	47%
Surface Evolver Bench	—	76.9%

Math Gemini 3.8 Flash leads

DeepSeek-V3.2-Exp: 41.7 (#87), Gemini 3.8 Flash: 65.3 (#28)

Math benchmarks
Benchmark	DeepSeek-V3.2-Exp	Gemini 3.8 Flash
OTIS Mock AIME 2024-2025	87.8%	98.9%
ProofBench	8%	48%
LMArena Math	1435	1528
FrontierMath (Tiers 1-3)	—	68.4%
FrontierMath Tier 4	—	22%
MathArena Final-Answer Competitions	57.7%	—
FrontierMath (Feb 2025 set)	22.1%	—
FrontierMath Tier 4 (v1)	2.1%	—

Knowledge Gemini 3.8 Flash leads

DeepSeek-V3.2-Exp: 51.7 (#66), Gemini 3.8 Flash: 74.8 (#2)

Knowledge benchmarks
Benchmark	DeepSeek-V3.2-Exp	Gemini 3.8 Flash
GPQA Diamond	83.4%	95.4%
LMArena Expert	1436	1524
Humanity's Last Exam	—	44.5%
SimpleQA Verified	—	69.7%
Vectara Hallucination Rate	5.3%	—

Multimodal Not comparable

DeepSeek-V3.2-Exp: —, Gemini 3.8 Flash: 40.7 (#45)

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

Multilingual Gemini 3.8 Flash leads

DeepSeek-V3.2-Exp: 52.2 (#90), Gemini 3.8 Flash: 58.0 (#5)

Multilingual benchmarks
Benchmark	DeepSeek-V3.2-Exp	Gemini 3.8 Flash
LMArena Non-English	1409	1491
LMArena Chinese	1461	1554
LMArena French	1433	1498
LMArena German	1440	1493
LMArena Japanese	1374	1502
LMArena Korean	1371	1459
LMArena Russian	1424	1515
LMArena Spanish	1440	1485

Instruction Following Gemini 3.8 Flash leads

DeepSeek-V3.2-Exp: 74.5 (#93), Gemini 3.8 Flash: 78.0 (#13)

Instruction Following benchmarks
Benchmark	DeepSeek-V3.2-Exp	Gemini 3.8 Flash
LMArena Instruction Following	1413	1490

Long Context DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 47.6 (#16), Gemini 3.8 Flash: 46.3 (#24)

Long Context benchmarks
Benchmark	DeepSeek-V3.2-Exp	Gemini 3.8 Flash
LMArena Longer Query	1428	1508
Fiction.LiveBench	83.3%	—
CL-bench	13.2%	—
CL-bench Life	9.5%	—

Writing & Preference Gemini 3.8 Flash leads

DeepSeek-V3.2-Exp: 62.4 (#77), Gemini 3.8 Flash: 72.2 (#15)

Writing & Preference benchmarks
Benchmark	DeepSeek-V3.2-Exp	Gemini 3.8 Flash
LMArena Text	1425	1499
LMArena Creative Writing	1403	1492
EQ-Bench Creative Writing	1515	1748
LMArena Multi-Turn	1427	1501

Frequently asked questions

Is DeepSeek-V3.2-Exp better than Gemini 3.8 Flash?

Which is cheaper, DeepSeek-V3.2-Exp or Gemini 3.8 Flash?

DeepSeek-V3.2-Exp is cheaper. It lists at $0.26 per million input tokens and $0.38 per million output tokens; Gemini 3.8 Flash lists at $0.75 and $3.75.

Is DeepSeek-V3.2-Exp or Gemini 3.8 Flash better for coding?

Gemini 3.8 Flash scores higher on coding benchmarks: 59.2 versus 46.5 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-V3.2-Exp and Gemini 3.8 Flash share?

34 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 49 scored results on Noometry and Gemini 3.8 Flash has 50.