OpenAI · automatic caching
o4 Mini-2025-04-16
o4 Mini-2025-04-16 on Zumik: live pricing, context, and caching, routable by id or alias through one OpenAI-compatible endpoint.
Specifications
At a glance.
| Provider | OpenAI |
| Family | o-series |
| Released | 2025-04 |
| License | Proprietary |
| Context window | 200K tokens |
| Max output | 100K tokens |
| Modalities | text, image |
| Tool calling | Yes |
| Reasoning mode | Yes |
| Caching | automatic |
| Batch discount | 50% off |
Measured by Zumik
What reuse looks like here.
Pricing, context, and capabilities for o4 Mini-2025-04-16 are live, but it is outside the flagship set Zumik benchmarks in depth, so measured reuse, capture, and warm TTFT are not shown yet. Run a workload estimate or route it by id to start collecting traces.
What you actually pay once caching works.
At a typical 55% prefix reuse, a million input tokens on o4 Mini-2025-04-16 effectively costs $0.65 instead of $1.10 - blending to roughly $1.58 with a 25% output share. Background work drops a further 50% on the batch tier.
Estimate it for your workloadRoute it directly by id, or let an alias pick it when it wins under policy.
Call it
Same OpenAI client, this model.
from openai import OpenAI
client = OpenAI(base_url="https://api.zumik.ai/v1", api_key="zk_live_...")
r = client.responses.create(
model="o4-mini-2025-04-16",
input="Draft a fix for the failing test.",
)
print(r.usage.input_tokens_cached) # confirm reuseFrequently asked
o4 Mini-2025-04-16, answered.
How much does o4 Mini-2025-04-16 cost?
o4 Mini-2025-04-16 is $1.10 per million input tokens and $4.40 per million output tokens through Zumik. Cache reads are $0.28 per million, a 75% discount on input.
What is o4 Mini-2025-04-16's context window?
o4 Mini-2025-04-16 supports a 200K-token context window with up to 100K output tokens.
Does o4 Mini-2025-04-16 support prompt caching?
Yes. OpenAI uses Automatic prefix caching caching. In the Zumik corpus, o4 Mini-2025-04-16 shows a median cache capture of 85% on agent workloads.
Run o4 Mini-2025-04-16 with reuse measured.
Point an OpenAI client at Zumik and see exactly how much of this model's input you are reusing.