The Known Good Updated 25 Jul 2026
DeepSeek: DeepSeek V3 vs OpenAI: gpt-oss-120b ·DeepSeek vs OpenAI

DeepSeek: DeepSeek V3 vs OpenAI: gpt-oss-120b — head-to-head comparison

DeepSeek: DeepSeek V3 profile OpenAI: gpt-oss-120b profile All comparisons
DeepSeek: DeepSeek V3 — DeepSeek
Quality
#44 of 61
44based on 3 of 6 evaluations
Known Good Index
Speed
#280 of 316
20.9
Output tokens per second
Price
#105 of 340
$0.35
Blended per 1M tokens
In $0.20 · Out $0.80
Cache hit price
#10 of 206
$0.003
per 1M tokens (−99%)
Context
#225 of 345
164k
tokens (~300 pages)
OpenAI: gpt-oss-120b — OpenAI
Quality
#33 of 61
61based on 3 of 6 evaluations
Known Good Index
Speed
#10 of 316
172.2
Output tokens per second
Price
#33 of 340
$0.07
Blended per 1M tokens
In $0.04 · Out $0.17
Cache hit price
#25 of 206
$0.015
per 1M tokens (−59%)
Context
#238 of 345
131k
tokens (~200 pages)
Summary

On the Known Good Index, DeepSeek: DeepSeek V3 scores 44 and OpenAI: gpt-oss-120b scores 61 — a 17-point lead for OpenAI: gpt-oss-120b. They share 2 published evaluations, with DeepSeek: DeepSeek V3 ahead on 0 and OpenAI: gpt-oss-120b ahead on 2.

Blended pricing is $0.35 per 1M tokens for DeepSeek: DeepSeek V3 and $0.07 for OpenAI: gpt-oss-120b — OpenAI: gpt-oss-120b is about 5.0× cheaper.

Throughput is 21 tokens per second for DeepSeek: DeepSeek V3 and 172 for OpenAI: gpt-oss-120b, with time to first token at 1.1s and 0.7s. DeepSeek: DeepSeek V3 accepts 164k tokens of context against 131k.

DeepSeek: DeepSeek V3 is open weights under Open weights (restricted use); OpenAI: gpt-oss-120b is open weights under Open weights (unrestricted). 18 tracked hosts serve one or both, and 2 serve both.

Generated from tracked data on 25 Jul 2026 · how these comparisons are computed →

At a glance
MetricDeepSeek: DeepSeek V3OpenAI: gpt-oss-120b
Known Good Index 44 based on 3 of 6 evaluations 61 based on 3 of 6 evaluations
Blended $/1M $0.35 $0.07
Output tok/s 21 172
Context window 164k 131k
Providers 3 17
Jump to section

Quality

Known Good Index and the capability indices, side by side

IndicesComponents
Index scores

Each index is min-max normalised across tracked models and scaled 0–100 from published evaluations. We do not run these evaluations.

MetricDeepSeek: DeepSeek V3OpenAI: gpt-oss-120b
Known Good Index 44 61
Mathematics Index 40 89
Science Index 53 77
Reasoning Index 52 —
Agentic Index — 18
Openness Index 60 60
The Known Good

Quality Breakdown

Every evaluation both models report a published score for

Evaluations both models report

Bar = DeepSeek: DeepSeek V3 · marker = OpenAI: gpt-oss-120b · scores ingested from Epoch AI (CC-BY 4.0)

GPQA Diamond57 / 76
MATH / AIME16 / 89

Right-hand figures read DeepSeek: DeepSeek V3 / OpenAI: gpt-oss-120b. DeepSeek: DeepSeek V3 leads on 0 of 2, OpenAI: gpt-oss-120b on 2.

The Known Good

Arena Elo

Human preference rating from blind pairwise votes

Arena ratings

Elo from blind pairwise votes, ingested from Arena. Higher is better.

MetricDeepSeek: DeepSeek V3OpenAI: gpt-oss-120b
Text Arena 1,333 1,366
Text Arena (style controlled) 1,358 1,352
The Known Good

Price & Cost

Published API pricing by token type

Per token typeCache discount
Published pricing

USD per 1M tokens. Blended is a 3:1 input:output mix. Live from OpenRouter, cross-checked against models.dev. Lower is better.

MetricDeepSeek: DeepSeek V3OpenAI: gpt-oss-120b
Input $/1M $0.20 $0.04
Output $/1M $0.80 $0.17
Blended $/1M $0.35 $0.07
Cache read $/1M $0.003 $0.015
The Known Good
We publish blended price rather than cost-per-task. Cost-per-task requires token counts from an evaluation suite we do not run. Blended price is computed from published rates and is fully reproducible — see the pricing methodology.

Context Window

Maximum input accepted and maximum output emitted

Context and output limits

Provider-declared limits, from metadata via models.dev and OpenRouter. Higher is better.

MetricDeepSeek: DeepSeek V3OpenAI: gpt-oss-120b
Context window 164k 131k
Max output tokens 16k 131k
The Known Good

Speed & Latency

Median throughput and time to first token across hosting providers

Output speedTime to first token
Throughput and latency

Medians across all tracked providers, derived from OpenRouter endpoint telemetry.

MetricDeepSeek: DeepSeek V3OpenAI: gpt-oss-120b
Output tokens/s 21 172
Time to first token 1.1s 0.7s
The Known Good

Providers

Which hosts serve each model, and at what price and speed

18 hosts · 2 serve both
Full provider benchmarks →
Provider DeepSeek: DeepSeek V3 blendedDeepSeek: DeepSeek V3 tok/s OpenAI: gpt-oss-120b blendedOpenAI: gpt-oss-120b tok/s
StreamLake $0.35 24 — —
DeepInfra $0.46 20 $0.07 246
Novita $0.62 14 $0.10 42
Cerebras — — $0.45 495
Groq — — $0.26 411
SambaNova — — $0.34 306
BaseTen — — $0.20 285
Nebius — — $0.26 281
Parasail — — $0.26 207
Mara — — $0.30 181
Amazon Bedrock · first party — — $0.26 166
Together — — $0.26 111
Google · first party — — $0.16 85
Phala — — $0.26 82
DigitalOcean — — $0.17 62
CoreWeave — — $0.07 41
Mancer 2 — — $0.17 41
SiliconFlow — — $0.15 26

Per-host pricing and telemetry from OpenRouter. A dash means that host does not serve that model.

The Known Good
Sources for this page
Pricing & providers — OpenRouter Metadata — models.dev Benchmarks — Epoch AI (CC-BY 4.0)Elo — Arena Full attribution →