Model comparison

DeepSeek-V3.2-Exp vs Grok 4.20 (Non-Reasoning)

Grok 4.20 (Non-Reasoning) is the stronger model overall, scoring 48.6 to 44.3 on the Noometry Index. DeepSeek-V3.2-Exp costs 5.4× less per token, which makes it the better buy when Grok 4.20 (Non-Reasoning)'s lead doesn't matter for your workload.

Last verified October 11, 2026. 36 shared benchmarks.

DeepSeek-V3.2-Exp

DeepSeek

44.3

Rank #78 Confirmed

Grok 4.20 (Non-Reasoning) xAI

48.6

Rank #54 Confirmed

Summary

They share 36 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 2 categories and Grok 4.20 (Non-Reasoning) in 7 categories; 8 gaps are clear of the uncertainty.
The widest gap is in reasoning, where Grok 4.20 (Non-Reasoning) leads 52.3 to 22.1.
The biggest single-benchmark swing is ARC-AGI-2: 4% for DeepSeek-V3.2-Exp and 65.1% for Grok 4.20 (Non-Reasoning).
DeepSeek-V3.2-Exp is cheaper at $0.26 / $0.38 per million input/output tokens, against $1.25 / $2.50 for Grok 4.20 (Non-Reasoning).
Grok 4.20 (Non-Reasoning) accepts more context: 1M tokens versus 164K.
DeepSeek-V3.2-Exp has downloadable open weights; the other is API-only.

Side by side

DeepSeek-V3.2-Exp and Grok 4.20 (Non-Reasoning) specifications
	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
Provider	DeepSeek	xAI
Noometry Index	44.3	48.6
Released	2025-09-29	2026-02-17
Weights	Open	Proprietary
Context window	164K	1M
Max output	66K	30K
Input $ / M tokens	$0.26	$1.25
Output $ / M tokens	$0.38	$2.50
Results tracked	49	46

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

Category by category

Coding DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 46.5 (#65), Grok 4.20 (Non-Reasoning): 42.1 (#112)

Coding benchmarks
Benchmark	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
LMArena WebDev	1362	1375
WeirdML	39.5%	52.3%
LMArena Coding	1454	1459
SWE-bench Verified (bash only)	70%	—
Aider Polyglot	74.2%	—
SWE-bench Multilingual	59%	—
SciCode	38.9%	—
ALE-Bench	—	1,150

Agentic & Tool Use Grok 4.20 (Non-Reasoning) leads

DeepSeek-V3.2-Exp: 32.7 (#59), Grok 4.20 (Non-Reasoning): 34.4 (#46)

Agentic & Tool Use benchmarks
Benchmark	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
Terminal-Bench	39.6%	57.3%
Vending-Bench 2	1,034	4,663
APEX-Agents	21.3%	—
Berkeley Function Calling Leaderboard	56.7%	—
TheAgentCompany	42.9%	—
τ²-bench Banking	—	18%
LMArena Search	—	1189

Reasoning Grok 4.20 (Non-Reasoning) leads

DeepSeek-V3.2-Exp: 22.1 (#208), Grok 4.20 (Non-Reasoning): 52.3 (#32)

Reasoning benchmarks
Benchmark	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
ARC-AGI-2	4%	65.1%
Kagi LLM Benchmark	52.2%	75%
NYT Connections (extended)	36.7%	85.4%
ARC-AGI-1	57%	89.5%
Chess Puzzles	14%	24%
Thematic Generalization	65%	63.8%
LMArena Hard Prompts	1434	1451
DTBench	87.7%	90.1%
LMCA	29.1%	38.7%
Epoch Capabilities Index	146.27	151.98
CritPt	2.9%	—
ForecastBench	—	61.4

Math Grok 4.20 (Non-Reasoning) leads

DeepSeek-V3.2-Exp: 41.7 (#87), Grok 4.20 (Non-Reasoning): 48.2 (#65)

Math benchmarks
Benchmark	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
OTIS Mock AIME 2024-2025	87.8%	92.2%
ProofBench	8%	14%
LMArena Math	1435	1455
FrontierMath (Tiers 1-3)	—	44.9%
FrontierMath Tier 4	—	17.1%
MathArena Final-Answer Competitions	57.7%	—
FrontierMath (Feb 2025 set)	22.1%	—
FrontierMath Tier 4 (v1)	2.1%	—

Knowledge Grok 4.20 (Non-Reasoning) leads

DeepSeek-V3.2-Exp: 51.7 (#66), Grok 4.20 (Non-Reasoning): 52.8 (#60)

Knowledge benchmarks
Benchmark	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
GPQA Diamond	83.4%	89.3%
LMArena Expert	1436	1439
SimpleQA Verified	—	30.2%
Vectara Hallucination Rate	5.3%	—

Multimodal Not comparable

DeepSeek-V3.2-Exp: —, Grok 4.20 (Non-Reasoning): 33.3 (#98)

Multimodal benchmarks
Benchmark	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
LMArena Vision	—	1263
Blueprint-Bench 2	—	0%
LMArena Document	—	1416

Multilingual Grok 4.20 (Non-Reasoning) leads

DeepSeek-V3.2-Exp: 52.2 (#90), Grok 4.20 (Non-Reasoning): 54.5 (#40)

Multilingual benchmarks
Benchmark	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
LMArena Non-English	1409	1441
LMArena Chinese	1461	1481
LMArena French	1433	1476
LMArena German	1440	1465
LMArena Japanese	1374	1449
LMArena Korean	1371	1417
LMArena Russian	1424	1458
LMArena Spanish	1440	1443

Instruction Following Too close to call

DeepSeek-V3.2-Exp: 74.5 (#93), Grok 4.20 (Non-Reasoning): 74.8 (#83)

Instruction Following benchmarks
Benchmark	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
LMArena Instruction Following	1413	1420

Long Context DeepSeek-V3.2-Exp leads

DeepSeek-V3.2-Exp: 47.6 (#16), Grok 4.20 (Non-Reasoning): 45.5 (#34)

Long Context benchmarks
Benchmark	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
CL-bench	13.2%	22.2%
CL-bench Life	9.5%	11.9%
LMArena Longer Query	1428	1437
Fiction.LiveBench	83.3%	—

Writing & Preference Grok 4.20 (Non-Reasoning) leads

DeepSeek-V3.2-Exp: 62.4 (#77), Grok 4.20 (Non-Reasoning): 65.7 (#44)

Writing & Preference benchmarks
Benchmark	DeepSeek-V3.2-Exp	Grok 4.20 (Non-Reasoning)
LMArena Text	1425	1451
LMArena Creative Writing	1403	1438
EQ-Bench Creative Writing	1515	1574
LMArena Multi-Turn	1427	1456

Frequently asked questions

Is DeepSeek-V3.2-Exp better than Grok 4.20 (Non-Reasoning)?

Which is cheaper, DeepSeek-V3.2-Exp or Grok 4.20 (Non-Reasoning)?

DeepSeek-V3.2-Exp is cheaper. It lists at $0.26 per million input tokens and $0.38 per million output tokens; Grok 4.20 (Non-Reasoning) lists at $1.25 and $2.50.

Is DeepSeek-V3.2-Exp or Grok 4.20 (Non-Reasoning) better for coding?

DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.5 versus 42.1 in the Noometry coding category.

Which has the bigger context window?

Grok 4.20 (Non-Reasoning) does, with 1M tokens against 164K.

How many benchmarks do DeepSeek-V3.2-Exp and Grok 4.20 (Non-Reasoning) share?

36 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 49 scored results on Noometry and Grok 4.20 (Non-Reasoning) has 46.