Gemini 2.5 Pro Computer Use Preview

Released
2025-10-01
Last refreshed
2026-10-17
Status
Researched 28d ago
DeprecatedProprietaryCommercial use: restrictedMultimodalRAGAgentsLong contextVisionJSON / Tool use

Gemini 2.5 Pro Computer Use Preview is a legacy integration reference; keep it only while you identify a current replacement.

Google DeepMind releases · 54 in the last 12 months · this family litChangelog →
Specifications
Released
2025-10-01
Lifecycle
Retired July 28, 2026
Context
1.048576M
Architecture
Decoder Only
Knowledge cutoff
2025-01
Specialization
agentic
Openness
Proprietary
License
ProprietaryRestricted
Weights
Not released
Code
Unknown
Training
Pretrained
Created by
Google DeepMind

Pioneering artificial intelligence research.

London, United Kingdom
Founded 2014
Website
Pricing
Output / 1M
$10.00
Input / 1M
$1.25

Cheapest of 1 route · GCP Vertex AI

Providers(1)
GCP Vertex AI
View 1 provider route

About

Specialized for browser control agents. $1.25/$10.00 (<=200K), $2.50/$15.00 (>200K). Available on AI Studio and Vertex AI; no free tier.

Providers(1)

Compare API pricing across 1 tracked provider route for input and output tokens, batch, and cached reads when available.

ProviderInput / 1MOutput / 1MRoute
GCP Vertex AI$1.25$10.00
Serverless

Deprecated provider routes(1)

ProviderStatusCatalog ID
Google AI Studio
DeprecatedRetired July 28, 2026
gemini-2.5-computer-use-preview-10-2025

Available via routers & gateways(13)

LiteLLM

Gateway

Open-source Python SDK and proxy server that unifies 100+ LLM APIs behind a single OpenAI-compatible interface, with load balancing, cost tracking, and configurable failover.

Free OSSGoogle AI StudioGCP Vertex AI

OpenRouter

Hybrid

Unified hybrid gateway to 500+ models from 80+ providers via a single OpenAI-compatible API, with an optional Auto Router that selects a model per prompt from market spend data.

Passthrough + feeGoogle AI StudioGCP Vertex AI

Portkey

Gateway

Production AI gateway routing to 1,600+ LLMs with failover, load balancing, caching, and guardrails; MIT-licensed open-source core is self-hostable. Now part of Palo Alto Networks as Prisma AIRS AI Gateway.

SubscriptionGoogle AI StudioGCP Vertex AI

AIRouter

Router

Commercial LLM router (branded "AI Router" on its site) that picks the best model per request via a Decision API or an OpenAI-compatible Chat Completions endpoint, with a private selection mode that routes on anonymized patterns; sold as BYOK subscription plans priced by routed tokens.

SubscriptionGoogle AI StudioGCP Vertex AI

Helicone

Gateway

Observability-first AI gateway with routing, caching, rate limiting, and request tracing; Apache-2.0 open-source platform with a managed hosted tier for logging and analytics.

SubscriptionGoogle AI StudioGCP Vertex AI

Kong AI Gateway

Gateway

Multi-LLM AI gateway built on Kong Gateway 3.x, adding semantic routing, load balancing, guardrails, and MCP traffic analytics as plugins over Kong's existing API management platform.

SubscriptionGoogle AI StudioGCP Vertex AI

Capabilities

VisionMultimodalJSON / Tool useStructured Outputs

Benchmark peer barsfor RAG

No task-mapped benchmark peers are available for this model yet.

Benchmark scores(2)

Scores are benchmark-specific and are direction-aware: the same numeric gap can mean very different outcomes across suites. Use the leaderboard context and this model's provider route to decide whether the winning margin is meaningful for your workload.
BenchmarkScoreVersionEvaluationSource
WebVoyager
88.9WebVoyager, 559 date-edited tasks (self-reported, official leaderboard)Observed 2026-10-17
Configuration: Gemini 2.5 Computer Use (self-reported)
Harness: Gemini Computer Use API reference implementation; Anchor browser env; human judging
Evaluator: Google DeepMind
Confidence: self-reported
Notes: Model card table (Gemini 2.5 Computer Use, 10-2025) + additional-info PDF https://storage.googleapis.com/deepmind-media/gemini/computer_use_eval_additional_info.pdf (published 2025-10-07). Self-reported: 559 of 643 tasks remaining after date edits/removal of infeasible tasks; Anchor environment, API reference implementation defaults (temperature=1), pass@1, majority vote of 3 human judges. Card warns self-reported WebVoyager results are hard to compare across labs. Same card also reports 79.9% measured by Browserbase on its harness (identical across APIs, same query set/websites/day, human evaluation). Retrieved 2026-10-17.
Source
AndroidWorld
69.7AndroidWorld, screenshot-onlyObserved 2026-10-17
Configuration: Gemini 2.5 Computer Use
Harness: Gemini Computer Use API with mobile custom functions; AndroidWorld default settings
Evaluator: Google DeepMind
Confidence: self-reported
Notes: Model card table (Gemini 2.5 Computer Use, 10-2025) + additional-info PDF https://storage.googleapis.com/deepmind-media/gemini/computer_use_eval_additional_info.pdf (published 2025-10-07). Measured by Google DeepMind: Computer Use API with mobile custom functions (open_app, long_press_at, go_home), Pixel 6 emulators Android 13, default max steps with fixed RNG, screenshot-only (no a11y tree). Retrieved 2026-10-17.
Source

Migration checks

No linked migration route is available for this model yet.

Compare Gemini 2.5 Pro Computer Use Preview with other models