The Known Good Updated 2 Sep 2026 Subscribe
Google: Gemma 3 27B vs Qwen: Qwen3.6 35B A3B ·Google vs Alibaba

Google: Gemma 3 27B vs Qwen: Qwen3.6 35B A3B — head-to-head comparison

Google: Gemma 3 27B profile Qwen: Qwen3.6 35B A3B profile All comparisons
Google: Gemma 3 27B — Google
Quality
#49 of 65
47based on 4 of 7 evaluations
Known Good Index
Speed
#301 of 339
20.6
Output tokens per second
Price
#71 of 439
$0.17
Blended per 1M tokens
In $0.08 · Out $0.45
Cache hit price
#103 of 322
$0.040
per 1M tokens (−50%)
Context
#371 of 473
131k
tokens (~200 pages)
Qwen: Qwen3.6 35B A3B — Alibaba
Quality
#37 of 65
66based on 3 of 7 evaluations
Known Good Index
Speed
#88 of 339
74.3
Output tokens per second
Price
#109 of 439
$0.30
Blended per 1M tokens
In $0.10 · Out $0.90
Cache hit price
#106 of 322
$0.050
per 1M tokens (−50%)
Context
#199 of 473
262k
tokens (~400 pages)
Summary

On the Known Good Index, Google: Gemma 3 27B scores 47 and Qwen: Qwen3.6 35B A3B scores 66 — a 19-point lead for Qwen: Qwen3.6 35B A3B. They share 2 published evaluations, with Google: Gemma 3 27B ahead on 0 and Qwen: Qwen3.6 35B A3B ahead on 2.

Blended pricing is $0.17 per 1M tokens for Google: Gemma 3 27B and $0.30 for Qwen: Qwen3.6 35B A3B — Google: Gemma 3 27B is about 1.7× cheaper.

Throughput is 21 tokens per second for Google: Gemma 3 27B and 74 for Qwen: Qwen3.6 35B A3B, with time to first token at 0.8s and 1.5s. Google: Gemma 3 27B accepts 131k tokens of context against 262k.

Google: Gemma 3 27B is open weights under Open weights (restricted use); Qwen: Qwen3.6 35B A3B is open weights under Open weights (unrestricted). 12 tracked hosts serve one or both, and 3 serve both.

Generated from tracked data on 3 Sep 2026 · how these comparisons are computed →

At a glance
MetricGoogle: Gemma 3 27BQwen: Qwen3.6 35B A3B
Known Good Index 47 based on 4 of 7 evaluations 66 based on 3 of 7 evaluations
Blended $/1M $0.17 $0.30
Output tok/s 21 74
Context window 131k 262k
Providers 5 10
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.

MetricGoogle: Gemma 3 27BQwen: Qwen3.6 35B A3B
Known Good Index 47 66
Mathematics Index 49 —
Reasoning Index 40 —
Openness Index 60 60
The Known Good

Quality Breakdown

Every evaluation both models report a published score for

Evaluations both models report

Bar = Google: Gemma 3 27B · marker = Qwen: Qwen3.6 35B A3B · scores ingested from Epoch AI (CC-BY 4.0)

GPQA Diamond48 / 84
Mock AIME 2024–2522 / 87

Right-hand figures read Google: Gemma 3 27B / Qwen: Qwen3.6 35B A3B. Google: Gemma 3 27B leads on 0 of 2, Qwen: Qwen3.6 35B A3B 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.

MetricGoogle: Gemma 3 27BQwen: Qwen3.6 35B A3B
Text Arena 1,358 —
Text Arena (style controlled) 1,365 —
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.

MetricGoogle: Gemma 3 27BQwen: Qwen3.6 35B A3B
Input $/1M $0.08 $0.10
Output $/1M $0.45 $0.90
Blended $/1M $0.17 $0.30
Cache read $/1M $0.040 $0.050
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.

MetricGoogle: Gemma 3 27BQwen: Qwen3.6 35B A3B
Context window 131k 262k
Max output tokens 118k 236k
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.

MetricGoogle: Gemma 3 27BQwen: Qwen3.6 35B A3B
Output tokens/s 21 74
Time to first token 0.8s 1.5s
The Known Good

Providers

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

12 hosts · 3 serve both
Full provider benchmarks →
Provider Google: Gemma 3 27B blendedGoogle: Gemma 3 27B tok/s Qwen: Qwen3.6 35B A3B blendedQwen: Qwen3.6 35B A3B tok/s
Phala $0.23 51 $0.47 90
DeepInfra $0.10 21 $0.31 25
Novita $0.14 15 — —
Nebius $0.15 11 — —
Parasail $0.17 10 $0.36 55
CoreWeave — — $0.50 137
Io Net — — $0.44 85
AkashML — — $0.30 74
AtlasCloud — — $0.42 58
Venice — — $0.33 52
SiliconFlow — — $0.55 51
Darkbloom — — $0.21 50

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 →