Model comparison

DeepSeek-V3.1 vs gpt-oss-20b

DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 32.5 on the Noometry Index. gpt-oss-20b costs 12× less per token, which makes it the better buy when DeepSeek-V3.1's lead doesn't matter for your workload.

Last verified October 11, 2026. 22 shared benchmarks.

DeepSeek-V3.1

DeepSeek

42.8

Rank #108 Confirmed

gpt-oss-20b OpenAI

32.5

Rank #255 Confirmed

Summary

They share 22 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 6 categories and gpt-oss-20b in 2 categories; 7 gaps are clear of the uncertainty.
The widest gap is in writing & preference, where DeepSeek-V3.1 leads 60.3 to 35.5.
The biggest single-benchmark swing is DTBench: 82.7% for DeepSeek-V3.1 and 68% for gpt-oss-20b.
gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.25 / $0.95 for DeepSeek-V3.1.
DeepSeek-V3.1 accepts more context: 164K tokens versus 131K.

Side by side

DeepSeek-V3.1 and gpt-oss-20b specifications
	DeepSeek-V3.1	gpt-oss-20b
Provider	DeepSeek	OpenAI
Noometry Index	42.8	32.5
Released	2025-08-21	2025-08-05
Weights	Open	Open
Context window	164K	131K
Max output	8K	16K
Input $ / M tokens	$0.25	$0.018
Output $ / M tokens	$0.95	$0.09
Results tracked	27	34

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

Category by category

Coding DeepSeek-V3.1 leads

DeepSeek-V3.1: 40.3 (#144), gpt-oss-20b: 37.6 (#192)

Coding benchmarks
Benchmark	DeepSeek-V3.1	gpt-oss-20b
WeirdML	38.4%	40.9%
LMArena Coding	1417	1306
SciCode	—	34.4%
ALE-Bench	—	566.05

Agentic & Tool Use Not comparable

DeepSeek-V3.1: —, gpt-oss-20b: 9.3 (#154)

Agentic & Tool Use benchmarks
Benchmark	DeepSeek-V3.1	gpt-oss-20b
Terminal-Bench	—	3.4%

Reasoning DeepSeek-V3.1 leads

DeepSeek-V3.1: 27.9 (#110), gpt-oss-20b: 19.3 (#261)

Reasoning benchmarks
Benchmark	DeepSeek-V3.1	gpt-oss-20b
Kagi LLM Benchmark	53.2%	53.2%
LMArena Hard Prompts	1417	1274
DTBench	82.7%	68%
LMCA	24.3%	14.5%
Epoch Capabilities Index	139.92	137.82
SimpleBench	40%	—
CritPt	—	1.4%
Chess Puzzles	—	4%
ForecastBench	58	—

Math Too close to call

DeepSeek-V3.1: 38.9 (#122), gpt-oss-20b: 39.4 (#103)

Math benchmarks
Benchmark	DeepSeek-V3.1	gpt-oss-20b
LMArena Math	1420	1317
OTIS Mock AIME 2024-2025	—	65.3%
Omni-MATH	—	56.5%

Knowledge DeepSeek-V3.1 leads

DeepSeek-V3.1: 43.7 (#90), gpt-oss-20b: 34.6 (#195)

Knowledge benchmarks
Benchmark	DeepSeek-V3.1	gpt-oss-20b
LMArena Expert	1405	1258
GPQA Diamond	—	60.8%
MMLU-Pro	—	74%
Vectara Hallucination Rate	5.5%	—
GPQA (HELM)	—	59.4%

Multilingual DeepSeek-V3.1 leads

DeepSeek-V3.1: 51.6 (#106), gpt-oss-20b: 42.2 (#197)

Multilingual benchmarks
Benchmark	DeepSeek-V3.1	gpt-oss-20b
LMArena Non-English	1400	1268
LMArena Chinese	1469	1314
LMArena German	1411	1255
LMArena Japanese	1378	1244
LMArena Korean	1337	1236
LMArena Russian	1405	1278
LMArena Spanish	1431	1267
LMArena French	1447	—

Instruction Following DeepSeek-V3.1 leads

DeepSeek-V3.1: 73.9 (#110), gpt-oss-20b: 61.8 (#240)

Instruction Following benchmarks
Benchmark	DeepSeek-V3.1	gpt-oss-20b
LMArena Instruction Following	1400	1236
IFEval	—	73.2%

Long Context gpt-oss-20b leads

DeepSeek-V3.1: 36.3 (#232), gpt-oss-20b: 37.9 (#209)

Long Context benchmarks
Benchmark	DeepSeek-V3.1	gpt-oss-20b
LMArena Longer Query	1422	1250
Fiction.LiveBench	52.8%	—

Writing & Preference DeepSeek-V3.1 leads

DeepSeek-V3.1: 60.3 (#98), gpt-oss-20b: 35.5 (#265)

Writing & Preference benchmarks
Benchmark	DeepSeek-V3.1	gpt-oss-20b
LMArena Text	1420	1287
LMArena Creative Writing	1401	1201
EQ-Bench Creative Writing	1436	666
LMArena Multi-Turn	1408	1268
WildBench	—	73.7%

Frequently asked questions

Is DeepSeek-V3.1 better than gpt-oss-20b?

Which is cheaper, DeepSeek-V3.1 or gpt-oss-20b?

gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; DeepSeek-V3.1 lists at $0.25 and $0.95.

Is DeepSeek-V3.1 or gpt-oss-20b better for coding?

DeepSeek-V3.1 scores higher on coding benchmarks: 40.3 versus 37.6 in the Noometry coding category.

Which has the bigger context window?

DeepSeek-V3.1 does, with 164K tokens against 131K.

How many benchmarks do DeepSeek-V3.1 and gpt-oss-20b share?

22 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and gpt-oss-20b has 34.