Compared on Nov 13, 2026.

deep-research-pro-preview-12-2025 shutdown

Q: What replaces deep-research-pro-preview-12-2025?

Google recommends deep-research-preview-04-2026, and also offers deep-research-max-preview-04-2026. Both were released on April 21, 2026 and have no shutdown date announced. You change the agent parameter of interactions.create.

Q: What does a SpecStream research report cost?

$0.10 for quick, $0.30 for standard and $1.00 for deep. The price is taken once, when the report is delivered, and a cancelled report is not charged. New accounts get $10 of credit, no card.

What to change.

Google lists deep-research-pro-preview-12-2025 under deprecated agents, with October 23, 2026 as its shutdown date. To stay on Gemini, set agent to deep-research-preview-04-2026 or deep-research-max-preview-04-2026 in interactions.create. Google estimates $1.00 to $7.00 a task there.

SpecStream is the live web data engine for teams and agents. POST /v1/research takes the same question and returns a written report with the page behind each cited sentence, at $0.10, $0.30 or $1.00 a report.

Get an API key

New accounts get $10 of credit, no card. It covers 100 quick reports or 33 standard reports.

SpecStream, the 12-2025 agent and the 04-2026 agents

	SpecStream	Gemini 12-2025 agent	Gemini 04-2026 agents
Status on Nov 13, 2026	`POST /v1/research` answers, on the same key as the Search API and the Extract API	Listed under deprecated agents. Released December 11, 2025, shutdown date October 23, 2026	Preview. Released April 21, 2026, with no shutdown date announced
The call	`POST https://api.specstream.ai/v1/research`, then `GET /v1/research/{id}`	`interactions.create` with `agent="deep-research-pro-preview-12-2025"`	The same call with `agent` set to `deep-research-preview-04-2026` or `deep-research-max-preview-04-2026`. `background=true` is required
What comes back	`report.content`, the text with `[1]` markers. `report.citations` gives each cited sentence its page, read time and passage. `report.sources` lists every page read	Text: cited reports	Text: cited reports, and images. Charts when `visualization` is `auto`
JSON in place of text	`"output": {"type": "json", "schema": {...}}` with your JSON Schema	Output is text	Google's guide says the Deep Research agent does not support structured outputs
What you control	`depth` (`quick`, `standard`, `deep`), `length` (`short`, `standard`, `long`), `include_domains` and `exclude_domains`, and a `webhook`	Powered by Gemini 3.1 Pro. Reads public web sources and private workspace data like Gmail and Drive	Adds collaborative planning, visualization, remote MCP servers and File Search. No custom function calling tools
Limits	`input` holds up to 15,000 characters	Input context window 1,048,576 tokens, output limit 65,536 tokens	Input context window 1,048,576 tokens, output limit 65,536 tokens. A task runs at most 60 minutes
How long a report takes	About one minute for `quick`, four minutes for `standard` and five minutes for `deep`	Optimized for long-running tasks	Google says most tasks finish within 20 minutes
Price	$0.10 quick, $0.30 standard, $1.00 deep, shown as `price_usd` in the reply to the `POST`. Taken once, when the report is delivered. A cancelled report is not charged	Tokens at Gemini list rates plus tool fees	Tokens at Gemini list rates plus tool fees. Google's estimate: $1.00 to $3.00 a task, and $3.00 to $7.00 on Max, at preview rates that can change

What changes, and why teams move the job to SpecStream

What Google ended, and what it left running

Google announced the deprecation in its release notes of October 8, 2026 and gave one date: October 23, 2026. Its deprecations page now lists deep-research-pro-preview-12-2025 under deprecated agents with that shutdown date and deep-research-preview-04-2026 as the recommended replacement. The same page says a model that is shut down is completely turned off and its endpoint is no longer available.

The two 04-2026 agents are still previews. Both were released on April 21, 2026, and Google lists no shutdown date for either. Google gave 15 days between the notice and the date for the 12-2025 agent, so plan for the same again.

On Gemini, the fix is one parameter

Google's migration note is one sentence: update the agent parameter in your interactions.create requests from deep-research-pro-preview-12-2025 to deep-research-preview-04-2026, or to deep-research-max-preview-04-2026. Google describes the 04-2026 agent as designed for speed and efficiency, and the Max agent as maximum comprehensiveness.

The new agents add collaborative planning, visualization, MCP servers and File Search. The output can now hold images, so code that expects text alone should read each content part by its type. Token limits are the same on both model pages.

A SpecStream report has a price before it starts

Google's figures are estimates. A typical Deep Research task uses about 80 searches, a Max task up to 160, and the agent decides how much to read. A SpecStream report costs $0.10 for quick, $0.30 for standard and $1.00 for deep. The reply to the POST shows the price as price_usd, the balance pays it once when the report is delivered, and a cancelled report costs nothing. See pricing.

Each cited sentence comes with its passage

report.content is the text with [1] markers. report.citations gives each cited sentence its page, the time it was read and the passage that supports it, and report.sources lists every page read in the run. Your reader checks a sentence against the passage without opening the whole page. The research report docs list every field.

JSON when your code needs fields

Google's guide says the Deep Research agent does not support structured outputs. On SpecStream, send "output": {"type": "json", "schema": {...}} with a JSON Schema and the report comes back in that shape. length sets a short, standard or long report, and include_domains and exclude_domains choose the sites.

Search and page reading on the same key

The SpecStream Search API scores 97.9 F-score on all 1,000 official SimpleQA Verified questions (978 correct), and 64.0 item F1 and 43.4 row F1 on all 200 official WideSearch tasks. The Extract API scores 0.908 F1 on WCXB. Scores and method are on the benchmarks page. The Search API costs $5 per 1,000 searches of up to 10 results, or $1 with speed=fast, and the Extract API costs $0.80 per 1,000 pages.

Other deep research APIs

Parallel Task API

SpecStream writes a report for $0.10, $0.30 or $1.00, a price you see before it starts. Parallel prices each run by processor, from $5 per 1,000 runs on lite to $2,400 per 1,000 on ultra8x. The ultra processor costs $300 per 1,000, or 30¢ a run, and takes 5 to 25 minutes. A SpecStream standard report costs the same 30¢ and takes about four minutes.

Tavily Research

One request takes 4 to 110 credits on mini and 15 to 250 on pro, at $0.008 a credit pay as you go, so 3¢ to $2 a request. The charge lands between the minimum and the maximum. A SpecStream report has one price for its depth, shown as price_usd before the work starts.

OpenAI deep research

OpenAI shut down o3-deep-research on July 23, 2026, and its deprecations page names gpt-5.6-sol as the replacement. SpecStream takes the same research prompt as input and returns the report with its citations for $0.10 to $1.00. The OpenAI deep research page has the details.

Gemini Deep Research with SpecStream's search

The 04-2026 agents take remote MCP servers as tools. Add SpecStream at https://api.specstream.ai/mcp with search_web and read_pages, and the agent searches and reads the live web through SpecStream at $5 per 1,000 searches of up to 10 results and $0.80 per 1,000 pages. The Gemini Deep Research page has the code.

How SpecStream scores

The same numbers as the benchmarks page, from the official tasks and scorers.

	SpecStream	Parallel	Exa	Tavily	Perplexity
Web researchWideSearch, GPT-5.6 Luna	60.9	45.5	53.0	47.9	47.0
Factual accuracySimpleQA Verified, GPT-5.6 Luna	96	94	91	94	94
Factual accuracySimpleQA Verified, GPT-5.6 Sol	99	97	91	92	95

How to move a deep-research-pro-preview-12-2025 call

1. Find the agent string. Search your code, config files and environment variables for deep-research-pro-preview-12-2025.

2. To stay on Gemini, change the agent value. Nothing else in the request changes. Keep background=True and poll the interaction by its id.

import time
from google import genai

client = genai.Client()

interaction = client.interactions.create(
    agent="deep-research-preview-04-2026",  # was "deep-research-pro-preview-12-2025"
    input="Research the history of Google TPUs.",
    background=True,
)

while (result := client.interactions.get(id=interaction.id)).status != "completed":
    time.sleep(10)
print(result.steps[-1].content[0].text)

3. To move the job to SpecStream, create a key under Agents & API in the workspace. New accounts get $10 of credit, no card. Send the prompt as input. The reply is 202 with the report's id and its price_usd.

curl -X POST https://api.specstream.ai/v1/research \
  -H "Authorization: Bearer $SPECSTREAM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"input": "Research the history of Google TPUs.", "depth": "standard"}'

4. Read it until status is completed, as you polled the interaction. report.content is the text, report.citations the cited sentences and report.sources every page read.

curl https://api.specstream.ai/v1/research/REPORT_ID \
  -H "Authorization: Bearer $SPECSTREAM_API_KEY"

5. If you streamed progress from Gemini, follow GET /v1/research/{id}/events (server-sent events, Last-Event-ID resumes) or pass "webhook": {"url": "https://..."}. Send an Idempotency-Key header on the POST so a retry returns the same report and is charged once.

An agent can run the same report over MCP at https://api.specstream.ai/mcp: research_price shows the price of each depth, start_research starts the report once the person who pays approves that price, and get_research reads it. The MCP guide has the setup for each client.

Get an API key

Frequently asked questions

Something we didn't answer? Talk to a human

When was deep-research-pro-preview-12-2025 shut down?

October 23, 2026 is the shutdown date on Google's deprecations page, where the agent is listed under deprecated agents. Google announced it in the release notes of October 8, 2026, 15 days before.

What replaces deep-research-pro-preview-12-2025?

Google recommends deep-research-preview-04-2026, and also offers deep-research-max-preview-04-2026. Both were released on April 21, 2026 and have no shutdown date announced. You change the agent parameter of interactions.create.

What is different in deep-research-preview-04-2026?

It adds four things: collaborative planning, visualization, MCP servers and File Search. Its output can hold images as well as cited reports. The token limits are the same: 1,048,576 input and 65,536 output.

How much does the 04-2026 Deep Research agent cost?

Google estimates $1.00 to $3.00 a task on Deep Research and $3.00 to $7.00 on Deep Research Max, at preview rates that can change. You pay tokens at Gemini list rates plus tool fees. A SpecStream research report costs $0.10, $0.30 or $1.00 by depth.

Will the 04-2026 preview agents be shut down too?

Google lists no shutdown date for either 04-2026 agent as of November 13, 2026. Both are previews, and the 12-2025 preview had 15 days between its notice and its date.

What does a SpecStream research report cost?

$0.10 for quick, $0.30 for standard and $1.00 for deep. The price is taken once, when the report is delivered, and a cancelled report is not charged. New accounts get $10 of credit, no card.

Does SpecStream have an API and an MCP server?

Yes, and new accounts get $10 of credit, no card. The report runs at POST https://api.specstream.ai/v1/research, and over MCP at https://api.specstream.ai/mcp with research_price, start_research and get_research.

Sources

SpecStream wrote this page. Each rival figure comes from that company's own page.

SpecStream pricingspecstream.ai/pricing, read Nov 13, 2026
SpecStream for agentsspecstream.ai/agents.md, read Nov 13, 2026
SpecStream benchmarksspecstream.ai/benchmarks, read Nov 13, 2026
Gemini deprecationsaistudio.google.com/docs/deprecations, read Nov 13, 2026
Gemini API release notesaistudio.google.com/docs/changelog, read Nov 13, 2026
Deep Research Pro preview model pageaistudio.google.com/docs/models/deep-research-pro-preview-12-2025, read Nov 13, 2026
Deep Research preview model pageaistudio.google.com/docs/models/deep-research-preview-04-2026, read Nov 13, 2026
Gemini Deep Research agentaistudio.google.com/docs/deep-research, read Nov 13, 2026
Gemini API pricingaistudio.google.com/docs/pricing, read Nov 13, 2026
Parallel pricingdocs.parallel.ai/getting-started/pricing.md, read Nov 13, 2026
Tavily API creditsdocs.tavily.com/documentation/api-credits, read Nov 13, 2026
OpenAI o3-deep-researchdevelopers.openai.com/api/docs/models/o3-deep-research, read Nov 13, 2026
OpenAI deprecationsdevelopers.openai.com/api/docs/deprecations, read Nov 13, 2026