OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Developer Tools & SDKs

Open Source Vizier

Open Source Vizier is a Python research interface and service for black-box optimization, including hyperparameter tuning. It proposes parameter settings for an objective whose results come from experiments, and provides user, developer and benchmarking APIs.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Open Source Vizier — OpenViST project profile in AI Developer Tools & SDKs.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you can evaluate a system at proposed parameter settings and report a metric, or when you need APIs for optimization research and benchmarking.

Who it is for

  • Machine-learning developers tuning model hyperparameters through repeated evaluations.
  • Researchers implementing or comparing black-box optimization algorithms.
  • Teams seeking a client-server interface for distributed, multi-client optimization studies.

Documented uses

  • Describe a search space and objective metric for an experiment-driven optimization task.
  • Request candidate parameter settings and submit measurements from an objective evaluation.
  • Build optimization algorithms using the developer API.
  • Benchmark and compare algorithms using objective functions and methods.
Before you choose

Requirements & limits

The project is open source under the Apache-2.0 license. The documented full-package requirement is Python 3.10 or later; client-only packages require Python 3.8 or later.

  • The minimal package installs the core service and client APIs; optional extras add JAX, algorithms, benchmarks or broader support.
  • The JAX extra is the documented quick start for the JAX-based Bayesian optimizer.
  • The user API, developer API and benchmarking API address different stages: using optimization, implementing methods and comparing methods.
  • The user supplies the objective evaluation and reports each measurement.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up a Python optimization study

Install the documented JAX extra and define a basic study that requests suggestions and submits objective measurements.

Before you start

  • Python 3.10 or later for the full package, or Python 3.8 or later for client-only packages.
  • An objective function that evaluates proposed parameter values and returns a metric.

Steps

  1. Install the documented quick-start package with the command below.

    Documented command
    pip install google-vizier[jax]

  2. Import clients and pyvizier from vizier.service.

  3. Define a study configuration with an algorithm, parameter search space and objective metric.

  4. Create a study, request suggestions, evaluate each suggestion with your objective function, and submit the resulting measurement.

Expected outcome

A configured study can iterate through suggested parameter settings and record the objective measurements returned by the user’s evaluation code.

Guide limitations

  • The example’s objective is a simple function; integrating a model or other system requires the user to provide its evaluation code.
  • The source describes an optional server for distributed multi-client settings but does not provide those deployment steps in the getting-started example.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
12 October 2026
Original website title
google/vizier · Official homepage title · retained in its source language
How we review sourcesSuggest a correction