Mistral 7B
Mistral AI · dense · 7.3B parameters · 32,768 context
Parameters
7.3B
Context Window
32K tokens
Architecture
Dense
Best GPU
A10G
Cheapest API
$0.07/M
Quality Score
56/100
Intelligence Brief
Mistral 7B is a 7.3B parameter DENSE model from Mistral AI, featuring Grouped Query Attention (GQA) with 32 layers and 4,096 hidden dimensions. With a 32,768 token context window, it supports code, math, multilingual. On standardized benchmarks, it achieves MMLU 62.5, HumanEval 32, GSM8K 52.2. The most cost-effective API deployment is via deepinfra at $0.07/M output tokens. For self-hosted inference, A10G delivers optimal throughput at $285/month.
Architecture Details
Memory Requirements
BF16 Weights
14.6 GB
FP8 Weights
7.3 GB
INT4 Weights
3.6 GB
GPU Compatibility Matrix
Mistral 7B is compatible with 95% of GPU configurations across 41 GPUs at 3 precision levels.
GPU Recommendations
BF16 · 1 GPU · vllm
100/100
score
Throughput
221.9 tok/s
Latency (ITL)
4.5ms
Est. TTFT
1ms
Cost/Month
$285
Cost/M Tokens
$0.49
BF16 · 1 GPU · vllm
100/100
score
Throughput
345.1 tok/s
Latency (ITL)
2.9ms
Est. TTFT
0ms
Cost/Month
$332
Cost/M Tokens
$0.37
BF16 · 1 GPU · vllm
100/100
score
Throughput
372.8 tok/s
Latency (ITL)
2.7ms
Est. TTFT
0ms
Cost/Month
$370
Cost/M Tokens
$0.38
Deployment Options
API Deployment
deepinfra
$0.07/M
output tokens
Single GPU
A10G
$285/mo
Min VRAM: 7 GB
Multi-GPU
RTX 3080 x2
436.1 tok/s
TP· $266/mo
API Pricing Comparison
| Provider | Input $/M | Output $/M | Badges |
|---|---|---|---|
| deepinfra | $0.07 | $0.07 | Cheapest |
| together | $0.20 | $0.20 |
Cost Analysis
| Provider | Input $/M | Output $/M | ~Monthly Cost |
|---|---|---|---|
| deepinfraBest Value | $0.07 | $0.07 | $1 |
| together | $0.20 | $0.20 | $2 |
Cost per 1,000 Requests
Short (500 tok)
$0.05
via deepinfra
Medium (2K tok)
$0.20
via deepinfra
Long (8K tok)
$0.70
via deepinfra
Performance Estimates
Throughput by GPU
VRAM Breakdown (A10G, BF16)
Precision Impact
bf16
14.6 GB
weights/GPU
~221.9 tok/s
fp8
7.3 GB
weights/GPU
int4
3.6 GB
weights/GPU
Quality Benchmarks
Capabilities
Features
Supported Frameworks
Supported Precisions
Where to Deploy Mistral 7B
Self-Hosted Infrastructure
Similar Models
Codestral Mamba 7B
7.3B params · hybrid
Quality: 50
from $0.60/M
Falcon Mamba 7B
7.27B params · hybrid
Quality: 50
from $0.15/M
DeepSeek Math 7B
7.24B params · dense
Quality: 50
Prometheus 2 7B
7.24B params · dense
Quality: 50
NV EmbedQA Mistral 7B
7.24B params · dense
Quality: 50
from $0.01/M
Frequently Asked Questions
How much VRAM does Mistral 7B need for inference?
Mistral 7B requires approximately 14.6 GB of VRAM at BF16 precision, 7.3 GB at FP8, or 3.6 GB at INT4 quantization. Additional VRAM is needed for KV-cache (131072 bytes per token) and activations (~1.00 GB).
What is the best GPU for Mistral 7B?
The top recommended GPU for Mistral 7B is the A10G using BF16 precision. It achieves approximately 221.9 tokens/sec at an estimated cost of $285/month ($0.49/M tokens). Score: 100/100.
How much does Mistral 7B inference cost?
Mistral 7B API inference starts from $0.07/M input tokens and $0.07/M output tokens. Self-hosted inference costs depend on your GPU configuration — use our ROI calculator for a detailed breakdown.