OpenAI · LLM API Modelretiring

gpt-3.5-turbo

gpt-3.5-turbo costs $0.50 per 1M input tokens and $1.50 per 1M output tokens.

Updated September 21, 2026
Input / 1M
$0.50
Output / 1M
$1.50
Cached Input
—
Context Window
16K
Standard Input
$0.50 / 1M
Base prompt input rate
Standard Output
$1.50 / 1M
Completion output rate
Prompt Caching -50%
50% Off
Cached prompt context
Batch API -50%
50% Off
Asynchronous 24h batch

Save up to 50% on OpenAI API token spend

Connect your OpenAI account — we'll optimize Batch API workflows and prompt caching efficiency.

Book a demo Connect OpenAI — free

OpenAI retires gpt-3.5-turbo on 23 Oct 2026. Successor: gpt-5.6-terra.

Flagship models

All rates listed in USD per 1 Million (1M) tokens unless noted

Standard
Pricing Metric / Parameter
Standard RATE
(per 1M)
Short context input$0.50
Short context cached input-
Short context cache writes-
Short context output$1.50
Long context input-
Long context cached input-
Long context cache writes-
Long context output-

Finetuning

All rates listed in USD per 1 Million (1M) tokens unless noted

StandardBatch
Pricing Metric / Parameter
Standard RATE
(per 1M)
Batch RATE
(per 1M)
Training$8.00$8.00
Input$3.00$1.50
Cached input--
Output$6.00$3.00

Source: OpenAI official pricing page, verified 2026-09-21. Prices in USD per 1M tokens.

AI / LLM
LLM API Cost Comparison

Compare OpenAI model costs against Anthropic & Gemini

Use our free LLM Cost Calculator to benchmark gpt-3.5-turbo token rates, Batch API discounts, and monthly estimated spend side-by-side with leading AI models.

Open LLM Cost Calculator →