Google's agentic research model for multi-step investigation, priced at $2.00/$12.00 per million tokens with text and image output.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
gemini deep research preview is Google's agentic research model for autonomous, multi-step investigation — the kind of task where a single-pass answer isn't good enough and the model needs to search, read, synthesize, and cite across several rounds before it's done. It shares its context window, output ceiling, and pricing with the Max Preview variant in Google's lineup, which makes the choice between the two less about raw capability and more about how much depth a given research task actually needs.
This model is built around planning and executing a sequence of steps rather than answering directly. Reasoning is on by default and function calling is supported, so it can call tools, read results, and decide what to do next inside a single research run instead of stopping after one pass.
At $2.00 per million input tokens and $12.00 per million output tokens, this sits well above Google's Flash-tier models on price — appropriate for a model doing the work of many chained calls internally rather than one lightweight request. The 1,048,576-token context window gives a research run room to carry forward source material and intermediate findings across a long investigation without truncating context partway through.
Input accepts text, image, video, audio, and pdf; output covers text and image, not text alone. That's useful for research output that benefits from an included chart or diagram rather than a purely textual report — a genuine differentiator from single-modality-output research tools.
Deep Research Max Preview carries the identical context window, pricing, and modality profile in this listing, with a capability score of 88.9 shared by both. The practical distinction is in the name: "Max" points at the highest-comprehensiveness setting Google offers, while this standard Deep Research Preview is the more general-purpose entry point for agentic research work. Teams unsure which to start with are better off defaulting here and only reaching for Max Preview once a task has proven it needs more exhaustive coverage.
Like its Max sibling, this is a preview release, not a model with long-term API stability guarantees yet. Pinning a specific version rather than assuming behavior stays fixed is worth doing if reliability matters for a production pipeline.
Use the recorded facts as a shortlist, then validate the model against representative inputs and production constraints.
Gemini Deep Research Preview is categorized for Any-to-any, Audio-text-to-text, Automatic speech recognition, Image-text-to-text, Image-to-image, Image-to-text, Text generation, Text-to-image, Text-to-text AI models, Video-text-to-text. Its current record accepts text, image, video, audio, pdf and produces text, image. Confirm file formats, preprocessing, and provider-specific request schemas before implementation.
The directory records 1.0M context and 65.5K maximum output. Listed token prices are $2.00 input and $12.00 output per 1M tokens. Treat missing values as unknown and recheck current commercial terms.
Recorded output speed is — and time to first token is —. Hosting route, region, prompt length, concurrency, and provider load can materially change both measurements.
This page separates sourced model facts from incomplete fields. Benchmark evidence is displayed only when its source is linked and verified. Before choosing Gemini Deep Research Preview, test task quality, tool reliability, safety behavior, data controls, rate limits, and total cost on the exact route you intend to use.
Related approved models in Text-to-image. Compare specifications and verify fit for your workload.
Google's highest-effort agentic research model, priced at $2.00/$12.00 per million tokens with a 1M-token context window and image output.
Image model for prompt-driven generation, editing, and visual design workflows
Image model for prompt-driven generation, editing, and visual design workflows
Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing
Nano Banana Pro for higher-fidelity image generation and design-heavy edits