Skip to content
Updated minutes ago
NVIDIA

Canary 1B

NVIDIA · dense · 1B parameters · 4,096 context

Quality
50.0

Parameters

1B

Context Window

4K tokens

Architecture

Dense

Best GPU

RTX 3080

Cheapest API

$0.04/M

Intelligence Brief

Canary 1B is a 1B parameter DENSE model from NVIDIA, featuring Multi-Head Attention (MHA) with 24 layers and 1,024 hidden dimensions. With a 4,096 token context window, it supports multilingual. The most cost-effective API deployment is via nvidia-nim at $0.04/M output tokens. For self-hosted inference, RTX 3080 delivers optimal throughput at $133/month.

Architecture Details

TypeDENSE
Total Parameters1B
Active Parameters1B
Layers24
Hidden Dimension1,024
Attention Heads8
KV Heads8
Head Dimension128
Vocab Size128,000

Memory Requirements

BF16 Weights

2.0 GB

FP8 Weights

1.0 GB

INT4 Weights

0.5 GB

KV-Cache per Token12288 bytes
Activation Estimate0.20 GB

GPU Compatibility Matrix

Canary 1B is compatible with 100% 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

RTX 3080optimal

BF16 · 1 GPU · vllm

90/100

score

Throughput

2.1K tok/s

Latency (ITL)

0.5ms

Est. TTFT

0ms

Cost/Month

$133

Cost/M Tokens

$0.02

Use this config →
RTX 4060optimal

BF16 · 1 GPU · vllm

90/100

score

Throughput

734.4 tok/s

Latency (ITL)

1.4ms

Est. TTFT

0ms

Cost/Month

$209

Cost/M Tokens

$0.11

Use this config →
RTX 3070optimal

BF16 · 1 GPU · vllm

90/100

score

Throughput

1.2K tok/s

Latency (ITL)

0.8ms

Est. TTFT

0ms

Cost/Month

$85

Cost/M Tokens

$0.03

Use this config →

Deployment Options

API

API Deployment

nvidia-nim

$0.04/M

output tokens

Self-Hosted

Single GPU

RTX 3080

$133/mo

Min VRAM: 1 GB

Scale

Multi-GPU

RTX 3080

2.1K tok/s

Best available config

API Pricing Comparison

ProviderInput $/MOutput $/MBadges
nvidia-nim$0.04$0.04
Cheapest

Cost Analysis

ProviderInput $/MOutput $/M~Monthly Cost
nvidia-nimBest Value$0.04$0.04$0

Cost per 1,000 Requests

Short (500 tok)

$0.03

via nvidia-nim

Medium (2K tok)

$0.11

via nvidia-nim

Long (8K tok)

$0.40

via nvidia-nim

Performance Estimates

Throughput by GPU

RTX 3080
2.1K tok/s
RTX 4060
734.4 tok/s
RTX 3070
1.2K tok/s

VRAM Breakdown (RTX 3080, BF16)

Weights
KV
Act
Weights 2.0 GBKV-Cache 1.6 GBActivations 1.6 GBOverhead 0.2 GB

Precision Impact

bf16

2.0 GB

weights/GPU

~2.1K tok/s

fp8

1.0 GB

weights/GPU

int4

0.5 GB

weights/GPU

Capabilities

Features

Tool Use Vision Code Math Reasoning Multilingual Structured Output

Supported Frameworks

tensorrt-llmvllm

Supported Precisions

BF16 (default)FP8INT4

Where to Deploy Canary 1B

Similar Models

Larger context, Lower qualityCompare →
Similar specsCompare →
Similar specsCompare →

Frequently Asked Questions

How much VRAM does Canary 1B need for inference?

Canary 1B requires approximately 2.0 GB of VRAM at BF16 precision, 1.0 GB at FP8, or 0.5 GB at INT4 quantization. Additional VRAM is needed for KV-cache (12288 bytes per token) and activations (~0.20 GB).

What is the best GPU for Canary 1B?

The top recommended GPU for Canary 1B is the RTX 3080 using BF16 precision. It achieves approximately 2.1K tokens/sec at an estimated cost of $133/month ($0.02/M tokens). Score: 90/100.

How much does Canary 1B inference cost?

Canary 1B API inference starts from $0.04/M input tokens and $0.04/M output tokens. Self-hosted inference costs depend on your GPU configuration — use our ROI calculator for a detailed breakdown.