When does o3-pro-2025-06-10 shut down on the OpenAI API?

On December 11, 2026, according to developers.openai.com.

What replaces o3-pro-2025-06-10?

The named replacement is gpt-5.6-sol (reasoning.mode: pro). It is the provider's named replacement, not our recommendation.

How does the price change?

On the OpenAI API, moving from o3-pro-2025-06-10 to gpt-5.6-sol goes from $20.00 to $4.00 per 1M input tokens and from $80.00 to $20.00 per 1M output tokens (list prices). On a mix of 3 input tokens to 1 output token, that is −77%.

OpenAI retires o3-pro-2025-06-10 on Dec 11, 2026: replacement and cost

OpenAI APIShutdown

Date

Dec 11, 2026

Model

o3-pro-2025-06-10

Named replacement

gpt-5.6-sol

Maker

OpenAI

Replacement note: reasoning.mode: pro.

The deprecation table names only the dated snapshot. OpenAI's o3-pro model page lists o3-pro-2025-06-10 as the snapshot behind the alias o3-pro; check which one your calls use.

Price before and after

Per 1M tokens	Before (o3-pro-2025-06-10)	After (gpt-5.6-sol)
Input	$20.00	$4.00
Output	$80.00	$20.00
Cached input	not published	$0.40
Blended (3 in : 1 out)	$35.00	$8.00 (−77%)

Per 1M tokens

Before (o3-pro-2025-06-10)

After (gpt-5.6-sol)

Input

$20.00

$4.00

Output

$80.00

$20.00

Cached input

not published

$0.40

Blended (3 in : 1 out)

$35.00

$8.00 (−77%)

List prices, USD. Blended = (3 × input + 1 × output) ÷ 4 — a stated reference mix, not your workload. Prices: developers.openai.com, checked Sep 29, 2026.

Before price is the o3-pro row on OpenAI's pricing page; OpenAI's o3-pro model page lists o3-pro-2025-06-10 as its snapshot. OpenAI's pricing page lists one per-token rate for gpt-5.6-sol and no separate rate for reasoning.mode: pro; pro mode can generate many more reasoning (output) tokens. The replacement price is GPT-5.6 Sol's promotional rate, which OpenAI says is available at least through November 21, 2026. OpenAI does not publish the rate after the promotion, so the forecast keeps the promotional rate; your cost may be higher once it ends.

Source

What it does to your bill

Prefilled with o3-pro. Change the tokens, or add the other models you use.

Example workload — replace with yours.

ModelInput tokens / mo (M)Output tokens / mo (M)% cached inputAfter Dec 11, 2026

Up to 8 models. Find your monthly token counts on the OpenAI usage dashboard or Claude Console usage.

Output tokens after a migration: 0%Newer models can produce more or fewer tokens for the same task. We assume the same count unless you change this.Monthly cloud bill (AWS/Azure/GCP), optional

From announced changes alone, this workload goes from $3,600/mo in Oct 2026 to $800/mo in Mar 2027 (−77.8%).

Heads-up: GPT-5.6 Sol: promotional price, available at least through Nov 21, 2026. The forecast keeps today's rate because the later rate is not published.

Month-by-month table

Month	Cost
Oct 2026	$3,600.00
Nov 2026	$3,600.00
Dec 2026	$1,703.23
Jan 2027	$800.00
Feb 2027	$800.00
Mar 2027	$800.00

Change in monthly cost per announced event
#	Date	Change	$/mo change	Source
1	Dec 11, 2026	o3-pro → gpt-5.6-sol (reasoning.mode: pro)	−$2,800.00	developers.openai.com

The replacement is the provider's named replacement, not our recommendation; quality and behavior can differ.

This link contains the numbers you entered.

Methodology and limits

Prices are list prices in USD per 1M tokens from the provider's pricing page; dates and named replacements come from the provider's deprecation page, quoted above with the date we last checked. The replacement is the provider's named replacement, not our recommendation — quality and behavior can differ. Discounts, committed-use or enterprise agreements, batch and regional pricing, and Azure OpenAI and Amazon Bedrock prices are not included. Dates marked "at least through" or "not before" are shown but never applied.Full methodology →