Skip to content
Updated minutes ago
Google

Gemma 2 27B

Google · dense · 27B parameters · 8,192 context

Quality
65.0

Parameters

27B

Context Window

8K tokens

Architecture

Dense

Best GPU

H20

Cheapest API

$0.27/M

Quality Score

65/100

Intelligence Brief

Gemma 2 27B is a 27B parameter DENSE model from Google, featuring Grouped Query Attention (GQA) with 46 layers and 4,608 hidden dimensions. With a 8,192 token context window, it supports structured output, code, math, multilingual. On standardized benchmarks, it achieves MMLU 75.2, HumanEval 45, GSM8K 80. The most cost-effective API deployment is via deepinfra at $0.27/M output tokens. For self-hosted inference, H20 delivers optimal throughput at $940/month.

Architecture Details

TypeDENSE
Total Parameters27B
Active Parameters27B
Layers46
Hidden Dimension4,608
Attention Heads32
KV Heads16
Head Dimension128
Vocab Size256,000

Memory Requirements

BF16 Weights

54.0 GB

FP8 Weights

27.0 GB

INT4 Weights

13.5 GB

KV-Cache per Token376832 bytes
Activation Estimate1.50 GB

GPU Compatibility Matrix

Gemma 2 27B is compatible with 62% of GPU configurations across 41 GPUs at 3 precision levels.

BF16 (Full)
FP8 (Half)
INT4 (Quarter)
Blackwell(7 GPUs)
B200 NVL (pair)360GB
B300288GB
B100 SXM192GB
GB200 NVL72 (per GPU)192GB
Hopper(7 GPUs)
H100 NVL 94GB (per GPU pair)188GB
H200 SXM141GB
H2096GB
GH20096GB
Ada Lovelace(11 GPUs)
L40S48GB
L4048GB
RTX 6000 Ada48GB
L2048GB
Ampere(16 GPUs)
A100 80GB SXM80GB
A100 80GB PCIe80GB
A1664GB
RTX A600048GB
Legend:No fitVery tightTightModerateGoodExcellent

GPU Recommendations

H20optimal

FP8 · 1 GPU · tensorrt-llm

100/100

score

Throughput

1.1K tok/s

Latency (ITL)

1.0ms

Est. TTFT

0ms

Cost/Month

$940

Cost/M Tokens

$0.34

Use this config →
H200 SXMoptimal

FP8 · 1 GPU · tensorrt-llm

95/100

score

Throughput

1.1K tok/s

Latency (ITL)

1.0ms

Est. TTFT

0ms

Cost/Month

$2553

Cost/M Tokens

$0.93

Use this config →
H100 SXMoptimal

FP8 · 1 GPU · tensorrt-llm

95/100

score

Throughput

1.0K tok/s

Latency (ITL)

1.0ms

Est. TTFT

0ms

Cost/Month

$1794

Cost/M Tokens

$0.66

Use this config →

Deployment Options

API

API Deployment

deepinfra

$0.27/M

output tokens

Self-Hosted

Single GPU

H20

$940/mo

Min VRAM: 27 GB

Scale

Multi-GPU

A100 40GB SXM x2

306.1 tok/s

TP· $1613/mo

API Pricing Comparison

ProviderInput $/MOutput $/MBadges
deepinfra$0.27$0.27
Cheapest
together$0.30$0.30

Cost Analysis

ProviderInput $/MOutput $/M~Monthly Cost
deepinfraBest Value$0.27$0.27$3
together$0.30$0.30$3

Cost per 1,000 Requests

Short (500 tok)

$0.19

via deepinfra

Medium (2K tok)

$0.76

via deepinfra

Long (8K tok)

$2.70

via deepinfra

Performance Estimates

Throughput by GPU

H20
1.1K tok/s
H200 SXM
1.1K tok/s
H100 SXM
1.0K tok/s

VRAM Breakdown (H20, FP8)

Weights
Act
Weights 27.0 GBKV-Cache 3.1 GBActivations 12.0 GBOverhead 1.4 GB

Precision Impact

bf16

54.0 GB

weights/GPU

fp8

27.0 GB

weights/GPU

~1.1K tok/s

int4

13.5 GB

weights/GPU

Quality Benchmarks

Average
67th percentile across all models
MMLU
75.2
Below Average (47th pctile)
HumanEval
45.0
Below Average (39th pctile)
GSM8K
80.0
Below Average (47th pctile)
MT-Bench
81.0
Bottom 25% (0th pctile)

Capabilities

Features

Tool Use Vision Code Math Reasoning Multilingual Structured Output

Supported Frameworks

vllmsglangtgitensorrt-llmollama

Supported Precisions

BF16 (default)FP8INT4

Where to Deploy Gemma 2 27B

Similar Models

Gemma 2 9B

9.2B params · dense

Quality: 68

from $0.10/M

Cheaper, Smaller modelCompare →

Gemma 3 27B

27B params · dense

Quality: 69

from $0.20/M

Larger contextCompare →
Larger context, Lower qualityCompare →
Larger contextCompare →
Larger context, Lower qualityCompare →

Frequently Asked Questions

How much VRAM does Gemma 2 27B need for inference?

Gemma 2 27B requires approximately 54.0 GB of VRAM at BF16 precision, 27.0 GB at FP8, or 13.5 GB at INT4 quantization. Additional VRAM is needed for KV-cache (376832 bytes per token) and activations (~1.50 GB).

What is the best GPU for Gemma 2 27B?

The top recommended GPU for Gemma 2 27B is the H20 using FP8 precision. It achieves approximately 1.1K tokens/sec at an estimated cost of $940/month ($0.34/M tokens). Score: 100/100.

How much does Gemma 2 27B inference cost?

Gemma 2 27B API inference starts from $0.27/M input tokens and $0.27/M output tokens. Self-hosted inference costs depend on your GPU configuration — use our ROI calculator for a detailed breakdown.