RecurrentGemma 2B
Google · dense · 2.7B parameters · 8,192 context
Parameters
2.7B
Context Window
8K tokens
Architecture
Dense
Best GPU
RTX 3080
Intelligence Brief
RecurrentGemma 2B is a 2.7B parameter DENSE model from Google, featuring Grouped Query Attention (GQA) with 26 layers and 2,560 hidden dimensions. With a 8,192 token context window, it supports multilingual. For self-hosted inference, RTX 3080 delivers optimal throughput at $133/month.
Architecture Details
Memory Requirements
BF16 Weights
5.4 GB
FP8 Weights
2.7 GB
INT4 Weights
1.4 GB
GPU Compatibility Matrix
RecurrentGemma 2B is compatible with 100% of GPU configurations across 41 GPUs at 3 precision levels.
GPU Recommendations
BF16 · 1 GPU · vllm
100/100
score
Throughput
684.0 tok/s
Latency (ITL)
1.5ms
Est. TTFT
0ms
Cost/Month
$133
Cost/M Tokens
$0.07
BF16 · 1 GPU · vllm
100/100
score
Throughput
244.8 tok/s
Latency (ITL)
4.1ms
Est. TTFT
1ms
Cost/Month
$209
Cost/M Tokens
$0.32
BF16 · 1 GPU · vllm
100/100
score
Throughput
403.2 tok/s
Latency (ITL)
2.5ms
Est. TTFT
0ms
Cost/Month
$85
Cost/M Tokens
$0.08
Deployment Options
API Deployment
No API pricing available
Single GPU
RTX 3080
$133/mo
Min VRAM: 3 GB
Multi-GPU
RTX 3080
684.0 tok/s
Best available config
API Pricing Comparison
No API pricing data available for this model.
Performance Estimates
Throughput by GPU
VRAM Breakdown (RTX 3080, BF16)
Precision Impact
bf16
5.4 GB
weights/GPU
~684.0 tok/s
fp8
2.7 GB
weights/GPU
int4
1.4 GB
weights/GPU
Capabilities
Features
Supported Frameworks
Supported Precisions
Where to Deploy RecurrentGemma 2B
Self-Hosted Infrastructure
Similar Models
Phi 2
2.7B params · dense
Quality: 50
Gemma 2 2B
2.6B params · dense
Quality: 44
PaLI-Gemma 3B
2.9B params · dense
Quality: 50
Gemma 1.1 2B
2.5B params · dense
Quality: 50
OpenELM 3B
3B params · dense
Quality: 50
Frequently Asked Questions
How much VRAM does RecurrentGemma 2B need for inference?
RecurrentGemma 2B requires approximately 5.4 GB of VRAM at BF16 precision, 2.7 GB at FP8, or 1.4 GB at INT4 quantization. Additional VRAM is needed for KV-cache (26624 bytes per token) and activations (~0.30 GB).
What is the best GPU for RecurrentGemma 2B?
The top recommended GPU for RecurrentGemma 2B is the RTX 3080 using BF16 precision. It achieves approximately 684.0 tokens/sec at an estimated cost of $133/month ($0.07/M tokens). Score: 100/100.
How much does RecurrentGemma 2B inference cost?
RecurrentGemma 2B inference costs vary by provider and GPU setup. Use our calculator for detailed cost estimates across all providers.