Last verified
RETIRED OCT 23 2026RETIRED REASONING200KTEXT MODELPROMPT CACHING

o4-mini API Pricing

Retired October 23, 2026: OpenAI shut o4-mini down on that date, and calls to it now fail; OpenAI names GPT-5.6 Terra as the substitute. This archive keeps the later mini reasoning price band. $1.1/M input, $4.4/M output, and $0.275/M cached input. It kept the same $1.10/$4.40 headline price as o3-mini while cutting cached input in half.

Archived input - per 1M tokens
$1.10/M
Source OpenAI deprecated
Archived output - per 1M tokens
$4.40/M
Use for invoice checks deprecated
Cached input
$0.275/M
Cache hit price discounted
Effective - agentic blend
$0.742/M
92/8 split - 82% cache
§ 01 / TERMINAL

o4-mini API cost calculator.

Archive calculator pre-loaded with the last o4-mini rates OpenAI published. Tweak spend, output mix, or cache hit rate to compare legacy bills with current replacements.

$ /mo
Workload split
Prompt cache hit rate
Tokens you can process
—
Words equivalent (English)
—
Effective rate
—
Open full calculator (all models · share URL · CSV) →
§ 02 / SCENARIOS

o4-mini cost on real workloads.

§ 03 / TOKENIZER

o4-mini token counter and cost.

What's a token →
Estimate · tiktoken-cl100k_base · ≈4 chars/token Auto-counts as you type

This is a chars-per-token approximation, not a real tokenizer. Actual tokens vary by language, code density, and tool-call overhead — counts are typically ±10–20% off for English prose, more for code or non-Latin scripts. For exact billing, use the vendor's official tokenizer.

Characters 288
Words 42
Tokens (estimated) 72 tokens
Cost as input · uncached $0.00008 USD
Cost as output · uncached $0.00032 USD
Cost as cached input $0.00002 USD
§ 04 / SHELF

o4-mini vs similar models.

All models →
Model Input /M Output /M Effective blended Context Best for
o4-mini Current $1.10 cache $0.275 $4.40 $0.742 agentic 92/8 200K Reasoning archive comparisons
GPT-5.5 $5.00 cache $0.50 $30.00 $3.61 +4.9× 1M Current OpenAI flagship
GPT-5.4 $2.50 cache $0.25 $15.00 $1.80 +2.4× 1.05M Current OpenAI daily driver
GPT-5.4 mini $0.75 cache $0.075 $4.50 $0.541 −27% 400K Current cheap OpenAI workloads
Claude Sonnet 5.5 $2.00 cache $0.10 $10.00 $1.29 +74% 1M Anthropic's current Sonnet
Gemini 3.8 Flash $0.75 cache $0.075 $3.75 $0.481 −35% · promo 1M Gemini long-context work
DeepSeek V4 Pro $1.32 cache $0.044 $3.96 $0.569 −23% 1M Low-cost reasoning and coding
§ 05 / DEEP LINKS

o4-mini cost by scenario.

All calculators →
§ 07 / BENCHMARKS

o4-mini benchmarks

All benchmarks →

o4-mini has published scores on 2 benchmarks we track; its best placing is #67 of 105 on ARC-AGI-2. All of these scores come from independent leaderboards, not from the lab.

Benchmark Score Rank Leader Source
ARC-AGI-2 Reasoning 6.1% ARC Prize verified, best result #67 of 105 GPT-6 Astra · 95% Official leaderboard ARC Prize Foundation
ARC-AGI-1 Reasoning 58.7% ARC Prize verified, best result #64 of 99 6 tied · 98.5% Official leaderboard ARC Prize Foundation

o4-mini API pricing FAQ.

Short answers for teams checking o4-mini pricing, status, and migration choices.

Q · 01 Can I still use o4-mini? +
No. OpenAI announced on April 22, 2026 that o4-mini would be shut down on October 23, 2026, and its deprecations page now says access was shut down on that date. The substitute OpenAI names is GPT-5.6 Terra. OpenAI's pricing page still prints this model's rates in its legacy table, but they no longer buy anything; the figures here are the last OpenAI published, kept for old invoices and migration math.
Q · 02 How much did o4-mini cost? +
o4-mini is archived at $1.1/M input and $4.4/M output.
Q · 03 Is cached-input pricing included? +
Yes. Cached input is shown at $0.275/M.
Q · 04 What replaced it? +
OpenAI's deprecations page names GPT-5.6 Terra as the substitute, with the shutdown on October 23, 2026; use the shelf to compare other current options.
Q · 05 Why keep this archive page? +
Archive pages preserve old price baselines, help explain invoice history, and capture migration search demand without treating old models as current recommendations.