简体中文
繁體中文
English
Français
Русский
日本語
Tiếng Việt
한국어
Deutsch
Español
Italiano
Português
العربية
ไทย
Bahasa Indonesia
Bahasa Melayu
Türkçe
Polski
Nederlands
Dansk
Norsk
Қазақшаgpt-5.4GPT-5.4 pricing (November 2026)
OpenAI lists GPT-5.4 at $2.50 per 1M input tokens and $15.00 per 1M output tokens.
On Crazyrouter the same model is $2.00 / $12.00 — 20% below list price.
That is $0.0025 / $0.015 per 1K tokens. GPT-5.4 is the #29 cheapest of the 40 chat models we track, by input price. 1.6× cheaper than GPT-5.6 Sol. Through Crazyrouter you save $0.50 on every 1M input tokens versus list.
Base-tier rates shown; longer requests can cost more. Monthly estimates assume every request stays in the first tier.
Call GPT-5.4 at $2.00 in / $12.00 out per 1M tokens. Same model, same API, no plan — pay per token.
This page lists token-based pricing for the developer API. Consumer app subscriptions are billed separately.
Count long requests before projecting monthly cost
The billing expression switches above 272,000 input tokens per request. A monthly total hides how often that boundary is crossed. The examples use current default-group rates and exclude cache; an account-specific discount can change the amount.
Illustrative token counts, not measured lengths of the request example. USD at current default-group rates, 500 output tokens, no cache hits or writes. Each request uses its applicable tier; character count is not token count.
| Input tokens | Output tokens | Estimated USD |
|---|---|---|
| 2,000 | 500 | $0.01000 |
| 272,001 | 500 | $1.09700 |
Source and scope
Gateway catalogue and billing schedule, reviewed 2026-10-11. This confirms published configuration, not end-to-end feature tests or vendor release claims.
Crazyrouter · Token pricing tiers
USD / 1M tokens. Base-tier rates shown; longer requests can cost more. Monthly estimates assume every request stays in the first tier.
| Input tokens per request | Input | Output | Cache read | Cache write | Cache write (1h) |
|---|---|---|---|---|---|
| ≤ 272,000 | $2.00 | $12.00 | $0.20 | — | — |
| > 272,000 | $4.00 | $18.00 | $0.40 | — | — |
OpenAI · Token pricing tiers
USD / 1M tokens. Base-tier rates shown; longer requests can cost more. Monthly estimates assume every request stays in the first tier.
| Input tokens per request | Input | Output | Cache read | Cache write | Cache write (1h) |
|---|---|---|---|---|---|
| ≤ 272,000 | $2.50 | $15.00 | $0.25 | — | — |
| > 272,000 | $5.00 | $22.50 | $0.50 | — | — |
GPT-5.4 price by provider
USD per 1M tokens. Cloud prices are the base tier in the default region; the linked source is authoritative.
Base-tier rates shown; longer requests can cost more. Monthly estimates assume every request stays in the first tier.
- OpenAI$2.50$15.00
- Azure$2.50$15.00
- AWS Bedrock$2.50$15.00
- Crazyrouter$2.00$12.00
| Provider | Input | Cached input | Output | Context | Source |
|---|---|---|---|---|---|
| OpenAI | $2.50 | $0.25 | $15.00 | 1.1M | OpenAI |
| Azure | $2.50 | $0.25 | $15.00 | 922K | Azure OpenAI |
| AWS Bedrock | $2.50 | $0.25 | $15.00 | 922K | AWS Bedrock |
| Crazyrouter | $2.00 | $0.20 | $12.00 | 1.1M | crazyrouter.com |
Google Vertex AI · OpenRouter: No verified quote
per 1M tokens · Snapshot Nov 13, 2026
What GPT-5.4 costs per month
Three typical workloads, priced on each provider from the table above.
Base-tier rates shown; longer requests can cost more. Monthly estimates assume every request stays in the first tier.
Prototype: 10M in / 2M out
Input 10M · Output 2M
$44.00
- OpenAI$55.00
- Azure$55.00
- AWS Bedrock$55.00
- Crazyrouter$44.00
Product: 100M in / 20M out
Input 100M · Output 20M
$440.00
- OpenAI$550.00
- Azure$550.00
- AWS Bedrock$550.00
- Crazyrouter$440.00
Scale: 1B in / 200M out
Input 1,000M · Output 200M
$4,400
- OpenAI$5,500
- Azure$5,500
- AWS Bedrock$5,500
- Crazyrouter$4,400
Price changes
- Sep 3, 2026 — OpenRouter started listing GPT-5.4 at $2.50 / $15.00 per 1M.
- Mar 10, 2026 — Azure started listing GPT-5.4 at $2.50 / $15.00 per 1M.
- Mar 5, 2026 — OpenAI started listing GPT-5.4 at $2.50 / $15.00 per 1M.
Alternatives to GPT-5.4
Same-family models and cheaper options for similar tasks.
Guides and comparisons
ComparisonClaude Sonnet 5 vs GPT-5.4: API Behavior, JSON Output, and Production Routing Tested
A production-focused Claude Sonnet 5 vs GPT-5.4 comparison using live Crazyrouter API evidence from July 2, 2026, including model availability, response IDs, JSON output behavior,
PricingGPT-5.4 Pricing Explained — Cached Input, Context Tiers, Batch API, and How to Save with Crazyrouter
Complete breakdown of GPT-5.4 API pricing — $2.50/$15 per MTok for short context, $5/$22.50 for long context, automatic caching at 10%, Batch API 50% off, and how Crazyrouter saves
PricingGPT-5.4 API Pricing in 2026: Compare Cost, Access, and Best Alternatives
Compare GPT-5.4 API pricing, payment methods, and gateway options. Learn the easiest way to access GPT-5.4 with one API key and lower friction.
Podcast
AI Dev Tools: The Crazyrouter Podcast
Short episodes for developers on AI APIs, gateways, cost control and coding tools.
Latest episodes
EP196 · Oct 29, 2026 · 6:53
AI API Security Posture — Turn Controls Into Continuous Evidence
A practical guide to AI API security posture: inventory exposure, verify controls continuously, protect keys and data, constrain tools, and turn security assumptions into evidence that operators can act on.
EP195 · Oct 28, 2026 · 7:21
AI API Policy Drift — Detect When Production Rules Stop Matching Intent
A practical guide to policy drift in AI APIs: version routing and safety rules, detect stale enforcement, compare intended and observed behavior, and restore control before small gaps become incidents.
GPT-5.4 pricing FAQ
How much does GPT-5.4 cost?
OpenAI charges $2.50 per 1M input tokens and $15.00 per 1M output tokens. Through Crazyrouter the same model is $2.00 / $12.00. Base-tier rates shown; longer requests can cost more. Monthly estimates assume every request stays in the first tier.
What is cached input pricing?
Tokens that repeat an earlier prompt prefix are billed at the cached rate ($0.25 per 1M when supported) instead of the full input price. Base-tier rates shown; longer requests can cost more. Monthly estimates assume every request stays in the first tier.
Is there a batch discount?
Yes. OpenAI offers batch pricing at $1.25 in / $7.50 out per 1M tokens for asynchronous jobs. Base-tier rates shown; longer requests can cost more. Monthly estimates assume every request stays in the first tier.
Is GPT-5.4 cheaper on Azure, Bedrock or Vertex?
Cloud providers usually match the list price and add regional or provisioned tiers; see the table above for the base rate on each. Crazyrouter applies its discount on top of list. Base-tier rates shown; longer requests can cost more. Monthly estimates assume every request stays in the first tier.
Sources and method
Provider reference prices use LiteLLM data and maintained entries. Provider links are references, not proof of daily official verification. Snapshot: Nov 13, 2026. Crazyrouter rates come from our billing table; confirm current prices, region and terms with the provider before purchase.
- OpenAI — openai.com/api/pricing
- Azure — azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service
- AWS Bedrock — aws.amazon.com/bedrock/pricing
- Crazyrouter — crazyrouter.com/pricing
- LiteLLM