GPT-5.4

OpenAIMCP Tool CallReasoning
openai/gpt-5.4Closed API

Agent-ready GPT for coding and computer-use workflows at a lower cost

Benchmark Pricing$2.50 / 1M in$15.00 / 1M out$0.25 cache read
Context: 1,050,000 tokens•Max Output: 128,000 tokens•Knowledge: 2025-08-31•Released: 2026-03-05

TECHNICAL SPECIFICATIONS

Context Window1,050,0001.1M tokens
Max Output128,000128K tokens
LicenseProprietaryCloud API only
Input Modalities
TextVisionPDF
Function / Tool CallingSupported (MCP Ready)
Structured OutputsGuaranteed Schema

MODEL CONTEXT PROTOCOL (MCP) READINESS

✓ GPT-5.4 natively supports OpenAI / Anthropic format tool calling and is fully compatible with MCP servers and agent loops.

MCP Context Window Recommendations

With a context capacity of 1,050,000 tokens, this model can handle extremely large MCP resource manifests, complete database schemas, and multiple active server tool definitions simultaneously without degradation.

STANDARDIZED BENCHMARKS & EVALS

SWE-Bench Proresolve rate
59.1
SWE-Atlas Codebase QnAscore
40.8
SWE-Atlas Codebase QnAscore
36.3
SWE-Atlas Refactoringscore
44.29
SWE-Atlas Test Writingscore
44.36
SWE-Atlas Test Writingscore
40
Artificial Analysis Coding Agent Indexaverage pass@1
53.6
SWE-Atlas Codebase QnApass@1
72.4
SWE-Bench Propass@1
18.4
Terminal-Benchpass@1
69.8
Artificial Analysis Coding Agent Indexaverage pass@1
52.2
SWE-Atlas Codebase QnApass@1
72.9
SWE-Bench Propass@1
18.9
Terminal-Benchpass@1
64.7
Terminal-Benchsuccess rate
75.1
GPQA Diamondaccuracy
92.8
Humanity's Last Examaccuracy
39.8
Humanity's Last Examaccuracy
52.1
OSWorld-Verifiedsuccess rate
75
BrowseCompaccuracy
82.7
GDPvalwins or ties
83
ARC-AGI-2accuracy
73.3
FrontierMathaccuracy
47.6
FrontierMathaccuracy
27.1
MMMU Proaccuracy
81.2

API CALL INVOCATION EXAMPLE

// Example: Calling GPT-5.4 via OpenAI SDK
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.OPENAI_API_KEY,
});

const response = await client.chat.completions.create({
  model: "openai/gpt-5.4",
  messages: [{ role: "user", content: "Hello! Can you inspect MCP tools?" }],
});

console.log(response.choices[0].message.content);

PRICING BY PROVIDER

15 providers
OpenAIgpt-5.4
$2.50 / $15.00$0.25 cache
Cortecsgpt-5.4
$2.90 / $15.45$0.24 cache
NanoGPTopenai/gpt-5.4-mini
$0.75 / $4.50$0.07 cache
SAP AI Coregpt-5.4
$2.50 / $15.00$0.25 cache
GitHub Copilotgpt-5.4-mini
$0.75 / $4.50$0.07 cache
CrossModelopenai/gpt-5.4-mini
$0.75 / $4.50$0.07 cache
Snowflake Cortexopenai-gpt-5.4
$2.50 / $15.00$0.25 cache
302.AIgpt-5.4-mini
$0.75 / $4.50$0.07 cache
Vivgridgpt-5.4-mini
$0.75 / $4.50$0.07 cache
LLM Gatewayopenai/gpt-5.4-pro
$30.00 / $180.00
Poeopenai/gpt-5.4-pro
$27.27 / $163.64
AI/ML APIopenai/gpt-5.4-pro
$41.26 / $247.57
Cloudflare AI Gatewayopenai/gpt-5.4-pro
$30.00 / $180.00
Venice AIopenai-gpt-54
$2.50 / $15.00$0.25 cache
Databricksdatabricks-gpt-5-4
$2.50 / $15.00$0.25 cache

Rates in USD per 1M tokens. Rates synchronized from models.dev public catalog.

Prompt Caching Active

This model supports prompt caching at $0.25 / 1M tokens, saving up to 90% on repetitive MCP tool definitions and system instructions.

MODEL RESOURCES