GPT-4o
OpenAI · moe · 200B parameters · 128,000 context
Parameters
200B
Context Window
125K tokens
Architecture
MoE
Best GPU
B200 SXM
Cheapest API
$10.00/M
Quality Score
85/100
Intelligence Brief
GPT-4o is a 200B parameter Mixture-of-Experts (16 experts, 2 active) model from OpenAI, featuring Grouped Query Attention (GQA) with 80 layers and 10,240 hidden dimensions. With a 128,000 token context window, it supports tools, vision, structured output, code, math, multilingual. On standardized benchmarks, it achieves MMLU 88.7, HumanEval 72, GSM8K 95.8. The most cost-effective API deployment is via openai at $10.00/M output tokens. For self-hosted inference, B200 SXM delivers optimal throughput at $17044/month.
Provider pricing
3 providers · canonical: openai| Provider | Input $/M | Output $/M ▲ | Notes |
|---|---|---|---|
| openaicanonical | $2.50 | $10.00 | cheapest input · cheapest output |
| openrouter | $2.50 | $10.00 | cheapest input · cheapest output |
| artificialanalysis | $2.50 | $10.00 | 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
5 suggestions- GPT-4 TurboGPT-4 · 50B$30.00/M out
- GPT-4o MiniGPT-4 · 8B$0.600/M out
- GPT-5 MiniGPT · 25B$2.00/M out
- GPT-5GPT · 90B$10.00/M out
- o1-minio1 · 70B$12.00/M out
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
400.0 GB
FP8 Weights
200.0 GB
INT4 Weights
100.0 GB
Fits on (single GPU) — most practical first
GPU Compatibility Matrix
GPT-4o is compatible with 8% of GPU configurations across 41 GPUs at 3 precision levels.
GPU Recommendations
BF16 · 4 GPUs · tensorrt-llm
93/100
score
Throughput
280.0 tok/s
Latency (ITL)
3.6ms
Est. TTFT
1ms
Cost/Month
$17044
Cost/M Tokens
$23.16
BF16 · 4 GPUs · tensorrt-llm
93/100
score
Throughput
280.0 tok/s
Latency (ITL)
3.6ms
Est. TTFT
1ms
Cost/Month
$17082
Cost/M Tokens
$23.21
BF16 · 2 GPUs · tensorrt-llm
93/100
score
Throughput
280.0 tok/s
Latency (ITL)
3.6ms
Est. TTFT
1ms
Cost/Month
$19929
Cost/M Tokens
$27.08
Deployment Options
API Deployment
openai
$10.00/M
output tokens
Single GPU
Requires multi-GPU setup (200 GB VRAM needed)
Multi-GPU
B200 SXM x4
280.0 tok/s
TP· $17044/mo
API Pricing Comparison
| Provider | Input $/M | Output $/M | Badges |
|---|---|---|---|
| openai | $2.50 | $10.00 | Cheapest |
| openrouter | $2.50 | $10.00 | |
| artificialanalysis | $2.50 | $10.00 |
Cost Analysis
| Provider | Input $/M | Output $/M | ~Monthly Cost |
|---|---|---|---|
| openaiBest Value | $2.50 | $10.00 | $63 |
| openrouter | $2.50 | $10.00 | $63 |
| artificialanalysis | $2.50 | $10.00 | $63 |
Cost per 1,000 Requests
Short (500 tok)
$3.25
via openai
Medium (2K tok)
$13.00
via openai
Long (8K tok)
$40.00
via openai
Performance Estimates
Throughput by GPU
VRAM Breakdown (B200 SXM, BF16)
Quality Benchmarks
Capabilities
Features
Supported Frameworks
Supported Precisions
Where to Deploy GPT-4o
Self-Hosted Infrastructure
Similar Models
GPT-4 Turbo
200B params · moe
Quality: 80
from $30.00/M
Yi-Lightning
200B params · moe
Quality: 50
from $0.99/M
Claude Opus 4
200B params · dense
Quality: 90
from $75.00/M
o1
200B params · moe
Quality: 93
from $60.00/M
GLM-5
200B params · dense
Quality: 51
from $0.00/M
Frequently Asked Questions
How much VRAM does GPT-4o need for inference?
GPT-4o requires approximately 400.0 GB of VRAM at BF16 precision, 200.0 GB at FP8, or 100.0 GB at INT4 quantization. Additional VRAM is needed for KV-cache (204800 bytes per token) and activations (~4.00 GB).
What is the best GPU for GPT-4o?
The top recommended GPU for GPT-4o is the B200 SXM (x4) using BF16 precision. It achieves approximately 280.0 tokens/sec at an estimated cost of $17044/month ($23.16/M tokens). Score: 93/100.
How much does GPT-4o inference cost?
GPT-4o API inference starts from $2.50/M input tokens and $10.00/M output tokens. Self-hosted inference costs depend on your GPU configuration — use our ROI calculator for a detailed breakdown.