Google · implicit caching
Deep Research Preview (Apr-21-2026)
Deep Research Preview (Apr-21-2026) on Zumik: live pricing, context, and caching, routable by id or alias through one OpenAI-compatible endpoint.
Specifications
At a glance.
| Provider | |
| Family | gemini |
| Released | 2026-04 |
| License | Proprietary |
| Context window | 131K tokens |
| Max output | 66K tokens |
| Modalities | text, image, video, audio, pdf |
| Tool calling | Yes |
| Reasoning mode | Yes |
| Caching | implicit |
| Batch discount | 50% off |
Measured by Zumik
What reuse looks like here.
Pricing, context, and capabilities for Deep Research Preview (Apr-21-2026) 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 Deep Research Preview (Apr-21-2026) effectively costs $1.01 instead of $2.00 - blending to roughly $3.76 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="deep-research-preview-04-2026",
input="Draft a fix for the failing test.",
)
print(r.usage.input_tokens_cached) # confirm reuseFrequently asked
Deep Research Preview (Apr-21-2026), answered.
How much does Deep Research Preview (Apr-21-2026) cost?
Deep Research Preview (Apr-21-2026) is $2.00 per million input tokens and $12.00 per million output tokens through Zumik. Cache reads are $0.20 per million, a 90% discount on input.
What is Deep Research Preview (Apr-21-2026)'s context window?
Deep Research Preview (Apr-21-2026) supports a 131K-token context window with up to 66K output tokens.
Does Deep Research Preview (Apr-21-2026) support prompt caching?
Yes. Google Gemini uses Implicit context caching caching. In the Zumik corpus, Deep Research Preview (Apr-21-2026) shows a median cache capture of 84% on agent workloads.
Run Deep Research Preview (Apr-21-2026) with reuse measured.
Point an OpenAI client at Zumik and see exactly how much of this model's input you are reusing.