🧠 OPENAI COMMAND GRID · GPT-5.5 · GPT-5.4 · O3 · GPT-4o mini · TOOL CALLING · API LATENCY · COST / TOKEN · LIVE STATUS 🧠 OPENAI COMMAND GRID · GPT-5.5 · GPT-5.4 · O3 · GPT-4o mini · TOOL CALLING · API LATENCY · COST / TOKEN · LIVE STATUS

OpenAI Grid

AI RUNTIME LIVE

A live command desk for comparing benchmark quality, cost, latency, and routing behavior across OpenAI model families.

OpenAI in AI Race All Labs

Top model

GPT-5.5

Flagship

96.7

MMLU

lat 124ms · ctx 1M

H/E 99.5 · Math 97.9

launched Mar 2026

value 263.8 · io 6x

edge 0.0 over o3

ops profile: enterprise-ready with tool-heavy routing

live request volume

11.2k

RPS now

throughput 1.4B

avg lat 179ms

avg MMLU 91.7 · avg HE 87.7

value lead GPT-4o mini

avg math 89.5 · avg ctx 582K

avg speed 100 · density lead GPT-4o mini

flow trend: sustained request growth with healthy variance

active users

3.0M

Tracked sessions

uptime 99.74% · queue 12

quality 97.9

quality leader GPT-4o mini

value spread 310.7 vs 310.7

behavior: adaptive routing across high-confidence work queues

best value model

GPT-4o mini

High-Efficiency

value 310.7

io 4x

ctx 128K · speed 170

quickest GPT-4o mini

launched Nov 2024 · HE champ GPT-5.5

best fit: heavy context + precision-first workloads

cheapest in / out

$0.15 / $0.6

per 1M tokens

ratio 4.0x

from GPT-4o mini

baseline lane: PoC, experimentation, and bulk jobs

best cost model

GPT-4o mini

High-Efficiency

density 0.01

price spread $0.45 / M

context champ GPT-5.5

quality 79.6 · fastest GPT-4o mini

deployment lens: best efficiency with acceptable response risk

market pulse

—

Updating...

Model board

MMLU + live metrics

Model-by-model runtime profiles with benchmark depth, operating cost, and live traffic behavior in one glance.

GPT-5.5

#1

Highest benchmark leader with strong tool routing and enterprise reliability.

Tier: Flagship

MMLU 96.7

Lat 124ms

Ctx 1M

Cost-eff 263.8x

H/E 99.5

Speed 95

Vision Yes

Tools Yes

Tags: Tool-first · SWE-bench · Long context

IO 6.0x idle 48

o3

#2

Highest-cost reasoning path for deep symbolic workloads.

Tier: Reasoning Specialist

MMLU 96.7

Lat 332ms

Ctx 200K

Cost-eff 188.0x

H/E 99.5

Speed 40

Vision Yes

Tools Yes

Tags: Math · Formal reasoning · Verifier

IO 4.0x idle 20

GPT-5.4

#3

Trusted production layer for large prompt batches and long context flows.

Tier: Legacy Flagship

MMLU 91.4

Lat 182ms

Ctx 1M

Cost-eff 232.8x

H/E 74.9

Speed 95

Vision Yes

Tools Yes

Tags: Balanced · Long context · Automation

IO 6.0x idle 48

GPT-4o mini

#4

Best economics for high-throughput tasks and short-turnaround generation.

Tier: High-Efficiency

MMLU 82

Lat 76ms

Ctx 128K

Cost-eff 310.7x

H/E 77

Speed 170

Vision Yes

Tools Yes

Tags: Streaming · Cost cap · High-throughput

IO 4.0x idle 85

Use-case routing map

Reasoning stack

GPT-5.5

Strongest mix of MMLU + latency stability for tool-heavy flows.

Long-context analysis

GPT-5.4

1M context with reliable behavior over large inputs.

Math and theorem checks

o3

Highest depth for complex symbolic and verification routines.

Scale jobs

GPT-4o mini

Lowest cost baseline when cost per token dominates.