GPT-5.4 vs GPT-5.2 API pricing
Updated
GPT-5.4: USD 2.50 input, USD 15 output per 1M tokens (OpenAI and 2 others). GPT-5.2: USD 1.75 input, USD 14 output per 1M tokens (OpenAI and 2 others). At 3 input tokens per output token, GPT-5.4 costs 17% more than GPT-5.2 (USD 5.625 vs USD 4.813 per 1M).
How do GPT-5.4 and GPT-5.2 differ on price?
- Cheapest input price: GPT-5.2 USD 1.75 against USD 2.50 per 1M tokens, so GPT-5.4 costs 1.43× as much (43% more).
- Cheapest output price: GPT-5.2 USD 14 against USD 15 per 1M tokens, so GPT-5.4 costs 1.07× as much (7% more).
- Among the 201 indexed models, GPT-5.4 ranks 183rd cheapest and GPT-5.2 181st on the 3:1 blend.
- No input:output mix changes the answer: GPT-5.2 is cheaper at every mix before cache discounts.
- Both list cached-input discounts: GPT-5.4 at 4 standard prices of 4, from USD 0.25 (OpenAI, 90% below that provider's input price); GPT-5.2 at 3 standard prices of 3, from USD 0.175 (OpenAI, 90% below that provider's input price).
- Only GPT-5.4 has a stated context window (1,050,000 tokens); no provider of GPT-5.2 states one on its pricing page.
- GPT-5.4 has 4 providers with a standard price, 1 more than GPT-5.2; 3 providers sell both (DigitalOcean, Microsoft Azure and OpenAI).
- On one account, DigitalOcean charges USD 5.625 for GPT-5.4 (its cheapest price) and USD 4.813 for GPT-5.2 (its cheapest price) at 3:1, tied with 2 other providers for the lowest combined price among the providers that sell both.
- Shopping around matters more for GPT-5.4: its dearest standard price is 1.1× its cheapest, against 1× for GPT-5.2.
- OpenAI's own API is one of 3 providers tied for the lowest standard price we found for GPT-5.4, USD 5.625 at 3:1.
- OpenAI's own API is one of 3 providers tied for the lowest standard price we found for GPT-5.2, USD 4.813 at 3:1.
- GPT-5.4 has a batch price at 2 providers, from USD 2.813 at 3:1 (OpenAI); GPT-5.2 has a batch price at 1 provider, from USD 2.406 at 3:1 (OpenAI).
- For the one-developer coding agent preset, the gap is USD 4,887 a year in GPT-5.2's favour (GPT-5.4 USD 2,408 a month, GPT-5.2 USD 2,000).
GPT-5.4 and GPT-5.2 prices side by side
USD per 1M tokens, standard tier. "Cheapest" is the lowest price at 3 input tokens to 1 output token; each line can come from a different provider.
| Price | GPT-5.4 | GPT-5.2 |
|---|---|---|
| Model page | GPT-5.4 OpenAI · 4 providers | GPT-5.2 OpenAI · 3 providers |
| Cheapest standard price, blended 3:1 | USD 5.625 USD 2.50 in / USD 15 out at OpenAI and 2 others OpenAI page checked 2026-11-14 | USD 4.813 lower USD 1.75 in / USD 14 out at OpenAI and 2 others OpenAI page checked 2026-11-14 |
| Cheapest input | USD 2.50 OpenAI source | USD 1.75 lower OpenAI source |
| Cheapest cached input | USD 0.25 OpenAI source | USD 0.175 lower OpenAI source |
| Cheapest output | USD 15 OpenAI source | USD 14 lower OpenAI source |
| Maker's own API | USD 2.50 / USD 15 OpenAI source | USD 1.75 / USD 14 OpenAI source |
| Median standard price | USD 2.50 / USD 15 input / output, 4 prices | USD 1.75 / USD 14 input / output, 3 prices |
| Context window | 1,050,000 tokens, largest stated | No provider states one on its pricing page. |
| Providers with a standard price | 4 | 3 |
- Lower marks the lower price on a line; tie, equal prices. Other tiers and free rows are on the model pages.
Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).
What do GPT-5.4 and GPT-5.2 cost for a real workload?
30 days at each model's cheapest standard price for the workload (presets of the cost calculator; cached input at the cached price where listed).
| Workload | GPT-5.4 | GPT-5.2 | Difference |
|---|---|---|---|
| Support chatbot 2,000 requests a day, 1,500 input tokens (60% cached), 300 output | USD 373.50 OpenAI | USD 324.45 OpenAI | USD 49.05 GPT-5.2 lower |
| RAG over documents 1,000 requests a day, 8,000 input tokens (30% cached), 500 output | USD 663.00 OpenAI | USD 516.60 OpenAI | USD 146.40 GPT-5.2 lower |
| Coding agent (one developer) 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning | USD 2,408 OpenAI | USD 2,000 OpenAI | USD 407.25 GPT-5.2 lower |
GPT-5.2 is cheaper on all three workloads. The widest gap is on RAG over documents: the dearer model costs 28% more than the cheaper one.
Does the cheaper model change with your input:output ratio?
GPT-5.2 is cheaper at every mix of input and output tokens, before cache discounts: its cheapest input price and its cheapest output price are both at or below GPT-5.4's, so the input:output ratio never changes the answer.
| Input : output tokens | GPT-5.4 | GPT-5.2 |
|---|---|---|
| 1 : 3 (long answers) | USD 11.88 OpenAI | USD 10.94 lower OpenAI |
| 1 : 1 | USD 8.75 OpenAI | USD 7.875 lower OpenAI |
| 3 : 1 (the site’s blend) | USD 5.625 OpenAI | USD 4.813 lower OpenAI |
| 10 : 1 (retrieval, long prompts) | USD 3.636 OpenAI | USD 2.864 lower OpenAI |
| 30 : 1 (agents, big contexts) | USD 2.903 OpenAI | USD 2.145 lower OpenAI |
USD per 1M tokens at the cheapest standard provider for each mix, before cache discounts. Thinking tokens bill as output.
What do GPT-5.4 and GPT-5.2 cost per character in other languages?
Not measured: the real cost by language dataset covers only tokenizers we could measure from published files, and GPT-5.4 and GPT-5.2 are not among them. Count tokens on your own text with the counters below.
Check it on your own traffic
- GPT-5.4 API pricing: all 4 providers and other tiers.
- GPT-5.2 API pricing: all 3 providers and other tiers.
- Cost calculator with GPT-5.4 and GPT-5.2 preloaded
- GPT token counter
- Prompt caching calculator
More comparisons: GPT-5.5 vs GPT-5.4 · GPT-5.2 vs GPT-5 · every GPT-5.4 comparison · every GPT-5.2 comparison · all 662 price comparisons
Questions
Is GPT-5.4 cheaper than GPT-5.2?
No. On a blend of 3 input tokens to 1 output token, GPT-5.2 costs USD 4.813 per 1M tokens and GPT-5.4 USD 5.625, at each model's cheapest standard price. For input alone, GPT-5.2 is cheaper (USD 1.75 against USD 2.50 per 1M); for output alone, GPT-5.2 is (USD 14 against USD 15). This compares price only, not quality.
What do GPT-5.4 and GPT-5.2 cost per 1M tokens?
GPT-5.4: from USD 2.50 input and USD 15 output at OpenAI and 2 others; OpenAI's own API charges USD 2.50 / USD 15; median of 4 standard prices USD 2.50 / USD 15. GPT-5.2: from USD 1.75 input and USD 14 output at OpenAI and 2 others; OpenAI's own API charges USD 1.75 / USD 14; median of 3 standard prices USD 1.75 / USD 14. Each price is read from the provider's own pricing page; the source is linked in the table.
Which providers sell both GPT-5.4 and GPT-5.2?
3 providers have a standard price for both: DigitalOcean (USD 5.625 and USD 4.813 at 3:1), Microsoft Azure (USD 5.625 and USD 4.813 at 3:1), OpenAI (USD 5.625 and USD 4.813 at 3:1). The first 3 tie for the lowest combined price. Buying each model from its own cheapest provider costs USD 5.625 and USD 4.813.
Which is cheaper for a coding agent or a chatbot, GPT-5.4 or GPT-5.2?
For a coding agent (1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning): USD 2,408 a month on GPT-5.4 (OpenAI) against USD 2,000 on GPT-5.2 (OpenAI), a difference of USD 407.25. For a support chatbot (2,000 requests a day, 1,500 input tokens (60% cached), 300 output): USD 373.50 a month on GPT-5.4 (OpenAI) against USD 324.45 on GPT-5.2 (OpenAI), a difference of USD 49.05. Each uses the cheapest standard price for that workload, cached input included. Try your own numbers.
At what input:output ratio does the cheaper of GPT-5.4 and GPT-5.2 change?
GPT-5.2 is cheaper at every mix of input and output tokens, before cache discounts: its cheapest input price and its cheapest output price are both at or below GPT-5.4's, so the input:output ratio never changes the answer.
Do GPT-5.4 and GPT-5.2 have cached input pricing?
GPT-5.4 from USD 0.25 per 1M cached input tokens (OpenAI; 4 of 4 standard prices list a discount). GPT-5.2 from USD 0.175 per 1M cached input tokens (OpenAI; 3 of 3 standard prices list a discount). Cached tokens are billed at the input price where no cache price is published.