#tHow Many Tokens?

← All models

GPT-5.2: token counter & pricing

OpenAI · exact (uses official tokenizer) · pricing as of 2026-05-31.

Updated · By · Methodology

Provider
OpenAI
API model ID
gpt-5.2
Context window
400,000 tokens
Input price
$1.75 per 1M tokens
Output price
$14.00 per 1M tokens
Tokenizer accuracy
exact (uses official tokenizer)
Pricing as of
2026-05-31

Open the counter to count tokens for GPT-5.2 in real time.

What is GPT-5.2?

GPT-5.2 is the second incremental update in the GPT-5 line, meaningfully better reasoning and tool use than GPT-5 / 5.1 at a modest price bump. $1.75 input / $14 output per 1M tokens.

It's the "still cheap but noticeably smarter" tier, the most commonly chosen GPT-5 model for agentic workloads as of mid-2026 per OpenAI's published usage trends.

How tokens are counted here

OpenAI's o200k_base tokenizer (same as the rest of the GPT-5 family). Computed in your browser via js-tiktoken. Exact.

Pricing notes

$1.75 input / $14 output per 1M. Cached input $0.175/M.

Same single-tier pricing as GPT-5.1, no long-context surcharge above 128K. The output rate is what jumps from 5.1 ($10) to 5.2 ($14), 40% more expensive on the output column where reasoning models tend to spend their bill.

When to use GPT-5.2

When not to use it:

Common questions

What's GPT-5.2 Pro?

GPT-5.2 Pro is the premium tier of the same line at $21/$168 per 1M, 12× the cost of base GPT-5.2 on input. Use only when you've measured GPT-5.2 falling short on your eval.

How does GPT-5.2 compare to o3?

Same input rate ($2 for o3 vs $1.75 for GPT-5.2). o3 is positioned for explicit reasoning tasks and bills you for invisible "reasoning tokens", actual o3 output cost is typically 5-15× the visible reply length. For most agent tasks GPT-5.2 is the cheaper effective choice.

Cached prompt savings?

Cached input at $0.175/M is 10% of standard. On loops with stable system prompts (most agents), this pays back fast.

Compare GPT-5.2 to other models