SOLAR 10.7B
Upstage · dense · 10.7B parameters · 4,096 context
Parameters
10.7B
Context Window
4K tokens
Architecture
Dense
Best GPU
A100 40GB SXM
Cheapest API
$0.30/M
Intelligence Brief
SOLAR 10.7B is a 10.7B parameter DENSE model from Upstage, featuring Grouped Query Attention (GQA) with 48 layers and 4,096 hidden dimensions. With a 4,096 token context window, it supports structured output, code, math, multilingual. The most cost-effective API deployment is via upstage at $0.30/M output tokens. For self-hosted inference, A100 40GB SXM delivers optimal throughput at $807/month.
Provider pricing
1 provider · canonical: upstage| Provider | Input $/M | Output $/M ▲ | Notes |
|---|---|---|---|
| upstagecanonical | $0.300 | $0.300 | cheapest input · cheapest output |
Prices update via the nightly pricing cron + admin approvals at /admin/ingest-queue. The leaderboard's Input/Output cells show the canonical rate above; this table shows the full spread.
Recent changes
Loading…
Related models
3 suggestions
Picks: same family first, then same vendor within ±2× params, then top tag-overlap matches. Price shown is the cheapest Output $/M across providers — the row's page shows the canonical anchor.
Architecture Details
Memory Requirements
BF16 Weights
21.4 GB
FP8 Weights
10.7 GB
INT4 Weights
5.3 GB
GPU Compatibility Matrix
SOLAR 10.7B is compatible with 89% of GPU configurations across 41 GPUs at 3 precision levels.
GPU Recommendations
BF16 · 1 GPU · vllm
95/100
score
Throughput
392.4 tok/s
Latency (ITL)
2.5ms
Est. TTFT
0ms
Cost/Month
$807
Cost/M Tokens
$0.78
BF16 · 1 GPU · vllm
95/100
score
Throughput
452.2 tok/s
Latency (ITL)
2.2ms
Est. TTFT
0ms
Cost/Month
$845
Cost/M Tokens
$0.71
BF16 · 1 GPU · vllm
95/100
score
Throughput
392.4 tok/s
Latency (ITL)
2.5ms
Est. TTFT
0ms
Cost/Month
$655
Cost/M Tokens
$0.64
Deployment Options
API Deployment
upstage
$0.30/M
output tokens
Single GPU
A100 40GB SXM
$807/mo
Min VRAM: 11 GB
Multi-GPU
A4000 x2
182.5 tok/s
TP· $323/mo
API Pricing Comparison
| Provider | Input $/M | Output $/M | Badges |
|---|---|---|---|
| upstage | $0.30 | $0.30 | Cheapest |
Cost Analysis
| Provider | Input $/M | Output $/M | ~Monthly Cost |
|---|---|---|---|
| upstageBest Value | $0.30 | $0.30 | $3 |
Cost per 1,000 Requests
Short (500 tok)
$0.21
via upstage
Medium (2K tok)
$0.84
via upstage
Long (8K tok)
$3.00
via upstage
Performance Estimates
Throughput by GPU
VRAM Breakdown (A100 40GB SXM, BF16)
Precision Impact
bf16
21.4 GB
weights/GPU
~392.4 tok/s
fp8
10.7 GB
weights/GPU
int4
5.3 GB
weights/GPU
Capabilities
Features
Supported Frameworks
Supported Precisions
Where to Deploy SOLAR 10.7B
Similar Models
mmE5-mllama-11b-instruct
10.6B params · dense
Quality: 50
Llama 3.2 11B Vision
11B params · dense
Quality: 9
from $0.18/M
Falcon 11B
11B params · dense
Quality: 50
Falcon 3 10B
10.3B params · dense
Quality: 50
Alpamayo 1.5-10B
10B params · dense
Quality: 70
Frequently Asked Questions
How much VRAM does SOLAR 10.7B need for inference?
SOLAR 10.7B requires approximately 21.4 GB of VRAM at BF16 precision, 10.7 GB at FP8, or 5.3 GB at INT4 quantization. Additional VRAM is needed for KV-cache (196608 bytes per token) and activations (~1.00 GB).
What is the best GPU for SOLAR 10.7B?
The top recommended GPU for SOLAR 10.7B is the A100 40GB SXM using BF16 precision. It achieves approximately 392.4 tokens/sec at an estimated cost of $807/month ($0.78/M tokens). Score: 95/100.
How much does SOLAR 10.7B inference cost?
SOLAR 10.7B API inference starts from $0.30/M input tokens and $0.30/M output tokens. Self-hosted inference costs depend on your GPU configuration — use our ROI calculator for a detailed breakdown.