The Known Good Updated 4 Sep 2026 Subscribe
DeepSeek: DeepSeek V3 vs DeepSeek: R1 Distill Llama 70B ·DeepSeek vs DeepSeek

DeepSeek: DeepSeek V3 vs DeepSeek: R1 Distill Llama 70B — head-to-head comparison

DeepSeek: DeepSeek V3 profile DeepSeek: R1 Distill Llama 70B profile All comparisons
DeepSeek: DeepSeek V3 — DeepSeek
Quality
#43 of 65
62based on 4 of 7 evaluations
Known Good Index
Speed
#302 of 338
21.8
Output tokens per second
Price
#152 of 442
$0.46
Blended per 1M tokens
In $0.32 · Out $0.89
Cache hit price
#60 of 324
$0.020
per 1M tokens (−94%)
Context
#335 of 476
164k
tokens (~300 pages)
DeepSeek: R1 Distill Llama 70B — DeepSeek
Quality
#41 of 65
63based on 4 of 7 evaluations
Known Good Index
Speed
#306 of 338
20.7
Output tokens per second
Price
#211 of 442
$0.80
Blended per 1M tokens
In $0.80 · Out $0.80
Cache hit price
unranked
—
per 1M tokens
Context
#467 of 476
8k
tokens (~13 pages)
Summary

On the Known Good Index, DeepSeek: DeepSeek V3 scores 62 and DeepSeek: R1 Distill Llama 70B scores 63, which puts them level. They share 10 published evaluations, with DeepSeek: DeepSeek V3 ahead on 7 and DeepSeek: R1 Distill Llama 70B ahead on 3.

Blended pricing is $0.46 per 1M tokens for DeepSeek: DeepSeek V3 and $0.80 for DeepSeek: R1 Distill Llama 70B — DeepSeek: DeepSeek V3 is about 1.7× cheaper.

Throughput is 22 tokens per second for DeepSeek: DeepSeek V3 and 21 for DeepSeek: R1 Distill Llama 70B, with time to first token at 1.2s and 1.0s. DeepSeek: DeepSeek V3 accepts 164k tokens of context against 8k.

DeepSeek: DeepSeek V3 is open weights under Open weights (unrestricted); DeepSeek: R1 Distill Llama 70B is open weights under Open weights (unrestricted). 3 tracked hosts serve one or both, and 1 serve both.

Generated from tracked data on 4 Sep 2026 · how these comparisons are computed →

At a glance
MetricDeepSeek: DeepSeek V3DeepSeek: R1 Distill Llama 70B
Known Good Index 62 based on 4 of 7 evaluations 63 based on 4 of 7 evaluations
Blended $/1M $0.46 $0.80
Output tok/s 22 21
Context window 164k 8k
Providers 3 1
Jump to section

Quality

Known Good Index and the capability indices, side by side

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 V3DeepSeek: R1 Distill Llama 70B
Known Good Index 62 63
Mathematics Index 57 71
Reasoning Index 62 59
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 = DeepSeek: R1 Distill Llama 70B · scores ingested from Epoch AI (CC-BY 4.0), LiveBench (Apache-2.0)

GPQA Diamond71 / 56
Mock AIME 2024–2549 / 51
MATH Level 565 / 90
LiveBench60 / 54
LiveBench Coding62 / 52
LiveBench Data Analysis61 / 56
LiveBench Instruction Following75 / 70
LiveBench Language47 / 24
LiveBench Mathematics61 / 58
LiveBench Reasoning57 / 68

Right-hand figures read DeepSeek: DeepSeek V3 / DeepSeek: R1 Distill Llama 70B. DeepSeek: DeepSeek V3 leads on 7 of 10, DeepSeek: R1 Distill Llama 70B on 3.

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 V3DeepSeek: R1 Distill Llama 70B
Text Arena 1,332 —
Text Arena (style controlled) 1,358 —
The Known Good

Price & Cost

Published API pricing by token type

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 V3DeepSeek: R1 Distill Llama 70B
Input $/1M $0.32 $0.80
Output $/1M $0.89 $0.80
Blended $/1M $0.46 $0.80
Cache read $/1M $0.020 —
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 V3DeepSeek: R1 Distill Llama 70B
Context window 164k 8k
Max output tokens 16k 7k
The Known Good

Speed & Latency

Mean throughput and time to first token across hosting providers, over successful measurements from the last 48 hours

Throughput and latency

Means across all tracked providers over successful measurements from the last 48 hours, derived from OpenRouter endpoint telemetry.

MetricDeepSeek: DeepSeek V3DeepSeek: R1 Distill Llama 70B
Output tokens/s 22 21
Time to first token 1.2s 1.0s
The Known Good

Providers

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

3 hosts · 1 serve both
Full provider benchmarks →
Provider DeepSeek: DeepSeek V3 blendedDeepSeek: DeepSeek V3 tok/s DeepSeek: R1 Distill Llama 70B blendedDeepSeek: R1 Distill Llama 70B tok/s
StreamLake $0.45 24 — —
Novita $0.62 17 $0.80 17
DeepInfra $0.46 11 — —

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)Benchmarks — LiveBench (Apache-2.0)Elo — Arena Full attribution →