Scitke

Tag: OpenAI

  • OpenAI Introduces New Tools for Businesses to Develop AI Agents

    OpenAI Introduces New Tools for Businesses to Develop AI Agents

    On Tuesday, OpenAI unveiled new tools aimed at helping developers and businesses create AI agents—automated systems capable of independently performing tasks—using the company’s AI models and frameworks.
    Credit: Depositphotos

    On Tuesday, OpenAI unveiled new tools aimed at helping developers and businesses create AI agents—automated systems capable of independently performing tasks—using the company’s AI models and frameworks.

    These tools are part of OpenAI’s new Responses API, which enables businesses to develop custom AI agents that can conduct web searches, scan internal files, and navigate websites, similar to OpenAI’s Operator product. The Responses API replaces the Assistants API, which OpenAI plans to phase out by mid-2026.

    The Hurdles of AI Autonomy

    Despite growing excitement around AI agents, the industry has struggled to clearly define or demonstrate their practical value. A recent example is Chinese startup Butterfly Effect’s Manus platform, which went viral but failed to meet many user expectations, highlighting the challenges of delivering truly autonomous AI.

    OpenAI aims to overcome these hurdles. “It’s pretty easy to demo your agent,” said Olivier Godemont, OpenAI’s API product head, in an interview with TechCrunch. “To scale an agent is pretty hard, and to get people to use it often is very hard.”

    Earlier this year, OpenAI introduced two AI agents in ChatGPT: Operator, which navigates websites, and Deep Research, which compiles research reports. While these tools showcased agentic capabilities, they lacked full autonomy.

    With the Responses API, OpenAI now offers businesses access to the core components behind its AI agents, allowing developers to build their own applications that could surpass current solutions in autonomy and usability.

    With the Responses API, developers can access the same AI models powering OpenAI’s ChatGPT Search tool: GPT-4o search and GPT-4o mini search. These models can browse the web for answers, citing sources as they generate responses.

    OpenAI claims these models are highly accurate. On its SimpleQA benchmark, which evaluates fact-based question answering, GPT-4o search scores 90%, while GPT-4o mini search scores 88%—outperforming the recently released GPT-4.5, which scores only 63%.

    Limitations of AI-Powered Search

    AI-powered search tools generally surpass traditional AI models in accuracy since they can look up information directly. However, they still struggle with certain challenges, including hallucinations and difficulties with short, navigational queries like “Lakers score today.” Reports also suggest that ChatGPT’s citations are not always reliable.

    The Responses API also features a file search utility that quickly retrieves information from a company’s databases. OpenAI assures that these files won’t be used for model training. Additionally, developers can integrate OpenAI’s Computer-Using Agent (CUA) model, which powers the Operator tool. This model generates mouse and keyboard actions, enabling automation for tasks like data entry and workflow management.

    Enterprises can choose to run the CUA model locally on their systems, as it is launching in a research preview. However, the consumer version available in Operator is limited to web-based actions.

    Despite these advancements, the Responses API doesn’t eliminate all technical hurdles in AI agents. GPT-4o search still provides incorrect answers 10% of the time, and OpenAI acknowledges that its CUA model is not yet fully reliable for automating operating system tasks, as it can make unintended errors.

    To support developers, OpenAI is also launching the Agents SDK, an open-source toolkit that helps integrate AI models with internal systems, implement safeguards, and monitor agent behavior for debugging and optimization. The SDK builds on OpenAI’s Swarm framework, released last year for multi-agent orchestration.

    OpenAI’s API product lead, Olivier Godemont, believes this year will be crucial in turning AI agents from demos into practical tools. CEO Sam Altman has similarly predicted that 2025 will be the year AI agents enter the workforce. Whether that vision materializes remains to be seen, but OpenAI’s latest releases signal a shift toward making AI agents more functional and impactful.

    Read the original article on: TechCrunch

    Read more: DeepSeek: A Complete Guide to the AI Chatbot App

  • One Year Later, OpenAI has Yet to Launch its Voice Cloning Tool

    One Year Later, OpenAI has Yet to Launch its Voice Cloning Tool

    Image Credits:Stefano Guidi / Getty Images

    In late March of last year, OpenAI introduced a “small-scale preview” of Voice Engine, an AI tool capable of cloning a person’s voice using just 15 seconds of audio. Nearly a year later, the tool remains in limited testing, with no clear timeline for a full release—or confirmation that it will be launched at all.

    OpenAI’s hesitation to roll out the technology broadly could stem from concerns over potential misuse or an effort to avoid regulatory scrutiny. The company has previously faced criticism for prioritizing rapid product releases over safety and for rushing to market ahead of competitors.

    In a statement to TechCrunch, an OpenAI spokesperson said the company continues to test Voice Engine with a select group of “trusted partners.”

    “We’re learning from how [our partners are] using the technology to enhance both its usefulness and safety,” the spokesperson explained. “So far, we’ve seen promising applications in speech therapy, language learning, customer support, video game characters, and AI avatars.”

    Delays Persist for OpenAI’s Voice Engine, Despite Its Advanced Speech Capabilities

    Voice Engine, the technology behind OpenAI’s text-to-speech API and ChatGPT’s Voice Mode, produces highly natural-sounding speech that closely mimics the original speaker. The tool converts written text into speech, constrained by certain content safeguards. However, its release has faced multiple delays and shifting timelines.

    In a June 2024 blog post, OpenAI explained that Voice Engine learns to predict the most likely sounds a speaker would produce based on a given text transcript, accounting for variations in voice, accent, and speaking style. This allows the model to generate not only spoken text but also “spoken utterances” that simulate how different speakers might read aloud.

    Originally known as Custom Voices, OpenAI had planned to integrate Voice Engine into its API on March 7, 2024, according to a draft blog post viewed by TechCrunch. The initial rollout was intended for up to 100 “trusted developers,” prioritizing those creating applications with a “social benefit” or demonstrating “innovative and responsible” uses of the technology. OpenAI had even trademarked the tool and set pricing: $15 per million characters for standard voices and $30 per million characters for HD-quality voices.

    At the last minute, OpenAI delayed the announcement. When Voice Engine was finally unveiled a few weeks later, it lacked a public sign-up option. Instead, OpenAI stated that access would remain restricted to a small group of around 10 developers the company had been collaborating with since late 2023.

    OpenAI Emphasizes Ethical Considerations in Voice Engine’s Limited Rollout

    “We hope to foster a discussion on the responsible use of synthetic voices and how society can adapt to these advancements,” OpenAI wrote in its Voice Engine announcement blog post in March 2024. “The insights from these conversations and small-scale tests will help us determine whether and how to deploy this technology on a larger scale.”

    OpenAI has been developing Voice Engine since 2022 and showcased it to global policymakers in 2023. The tool, accessible to select partners like Livox, offers high-quality voice cloning but remains online-only, limiting its usability for some.

    Livox CEO Carlos Pereira praised its multilingual capabilities but hopes for an offline version. OpenAI has not provided updates on a broader launch or pricing, and Livox currently uses it for free.

    A key reason for Voice Engine’s delay is the risk of misuse, especially during elections. OpenAI has added safeguards like watermarking and requires explicit consent for voice cloning, but enforcing these policies at scale remains a challenge.

    The company is also exploring voice authentication and restrictions on cloning public figures. With AI voice scams on the rise, Voice Engine’s future remains uncertain, making it one of OpenAI’s longest-running limited previews.

    Read the original article on: TechCrunch

    Read more: Apple Introduces AI-Generated App Review Summaries in iOS 18.4

  • ChatGPT Pricing: A Complete Guide to OpenAI’s Subscription Plans and Costs

    ChatGPT Pricing: A Complete Guide to OpenAI’s Subscription Plans and Costs

    OpenAI’s ChatGPT platform continues to evolve, introducing new features to enhance user experience. Its memory function personalizes interactions by saving preferences, while an improved voice mode enables near real-time conversations. Additionally, the GPT Store offers a marketplace for AI-powered applications and services.
    Image Credits:Didem Mente/Anadolu Agency / Getty Images

    OpenAI’s ChatGPT platform continues to evolve, introducing new features to enhance user experience. Its memory function personalizes interactions by saving preferences, while an improved voice mode enables near real-time conversations. Additionally, the GPT Store offers a marketplace for AI-powered applications and services.

    If you’re wondering about ChatGPT’s pricing, the answer isn’t straightforward. OpenAI provides a range of plans, both free and paid, catering to individuals, nonprofits, businesses, educational institutions, and enterprises.

    To help you navigate the available subscription options, we’ve compiled a comprehensive guide on ChatGPT pricing, which we’ll update as new plans emerge.

    In the past, the free version of ChatGPT had significant limitations. However, OpenAI has since expanded its capabilities with enhanced generative AI models and new features.

    Expanded Features for Free Users

    Free ChatGPT users now have access to OpenAI’s GPT-4o mini model, web-augmented responses, the GPT Store, and the ability to upload files and photos for analysis. They also get limited access to advanced features such as Advanced Voice Mode, GPT-4o, and o3-mini. Additionally, users can save chat preferences as “memories” and utilize advanced data analysis, which allows ChatGPT to process files like spreadsheets and PDFs.

    Despite these improvements, the free plan has drawbacks, including daily usage limits on GPT-4o and file uploads, which vary based on demand. It also excludes certain advanced features, which we’ll explore in more detail below.

    For individuals seeking a more advanced ChatGPT experience, the ChatGPT Plus plan is available for $20 per month.

    ChatGPT Plus provides greater capacity than the free version, allowing users to send up to 80 messages to GPT-4o every three hours and unlimited messages to GPT-4o mini. Subscribers also gain access to OpenAI’s reasoning models, including o3-mini, o1-preview, and o1-mini.

    Additionally, ChatGPT Plus includes multimodal features such as Advanced Voice Mode with video and screen sharing, though daily limits may apply. Subscribers also receive limited access to emerging tools, including OpenAI’s deep research agent and Sora’s video generation.

    Advanced Data Analysis with GPT-4o

    The plan also offers an enhanced data analysis feature powered by GPT-4o, enabling users to generate interactive charts and tables from datasets. Files for analysis can be uploaded directly from Google Drive, Microsoft OneDrive, or local devices.

    For those seeking near-unlimited access to OpenAI’s products and early access to new features, ChatGPT Pro is available for $200 per month.

    Subscribers to ChatGPT Pro receive unlimited access to reasoning models, GPT-4o, and Advanced Voice Mode. This plan also includes 120 deep research queries per month and access to o1 Pro mode, which utilizes more computing power than the o1 version available in ChatGPT Plus.

    Additionally, Pro users can access OpenAI’s web-browsing agent, Operator, and generate more videos with Sora.

    OpenAI typically rolls out new features to ChatGPT Pro users first, and they receive priority access to existing features, such as GPT-4o, during peak usage times.

    If you own a small business or organization and need multiple ChatGPT licenses with collaboration features, ChatGPT Team could be the right choice. It costs $30 per user per month or $25 per user per month when billed annually for up to 149 users.

    ChatGPT Team: Collaborative AI for Organizations

    ChatGPT Team includes a dedicated workspace and admin tools for team management. All members gain access to OpenAI’s latest models, along with tools for analyzing, editing, and extracting information from files. Additionally, the plan allows teams to create and share custom apps—similar to those in the GPT Store—built on OpenAI models. These apps can be tailored for specific departments, use cases, or fine-tuned with team-specific data.

    OpenAI offers ChatGPT Enterprise for organizations needing more than 149 ChatGPT licenses. While pricing isn’t public, reports estimate $60 per user per month, with a 150-user minimum and a 12-month contract.

    This plan includes enterprise-grade privacy, advanced data analysis, improved performance, and customization options. It features a dedicated workspace, admin tools for user management, and integrations like SSO and domain verification.

    Companies can create internal workflows and bots with shareable conversation templates and receive API credits for custom ChatGPT solutions. Enterprise users also get priority access to models, dedicated support, and eligibility for Business Associate Agreements (BAAs) for handling private health data.

    OpenAI for Nonprofits is OpenAI’s initial step into nonprofit-focused tech solutions, offering discounts rather than a separate ChatGPT plan.

    Eligible nonprofits can subscribe to ChatGPT Team for $20 per user per month, while larger organizations receive a 50% discount on ChatGPT Enterprise, reducing the cost to around $30 per user.

    However, eligibility is limited. While nonprofits worldwide can apply, OpenAI currently excludes academic, medical, religious, and governmental institutions from the program.

    Read the original article on: TechCrunch

    Read more: ChatGPT’s Advanced Research: Can It Replace a Human Expert?

  • The Genius Method That Made OpenAI The Hottest Startup in Tech

    The Genius Method That Made OpenAI The Hottest Startup in Tech

    The hottest startup in Silicon Valley today is OpenAI, the Microsoft-backed developer of ChatGPT, a much-hyped chatbot that could write a poem, university essay, and even a line of software code.

    Tesla tycoon Elon Musk was an early investor in OpenAI. Microsoft, reportedly, provided approximately a first investment of $1 billion to $10 billion in an goal to challenge Google’s world-dominating search engine.

    If agreed, the money injection by the Windows maker should value OpenAI at a impressive $29 billion, making it a rare tech-world success when big players like Amazon, Meta, and Twitter are reducing costs and laying off staff.

    According to Dan Ives, analyst of Wedbush Securities, Microsoft is clearly being aggressive on this front and not going to be left behind on what can be a potential game-changing AI investment.

    Before the launch of ChatGPT, OpenAI had amazed tech geeks with Dall-E 2, software that produces digital images with simple instructions.

    Microsoft, that makes no secret of its AI ambitions, has integrated Dall-E 2 into numerous of its applications. Currently, according to a report in Bloomberg, the tech giant wishes to graft ChatGPT to its Bing search engine to take on Google.

    Since ChatGPT was introduced in November, this chatbot’s prowess has aroused internet users’ curiosity and fascination.

    It can formulate detailed and human-like answers on a wide variety of subjects in a few seconds, raising fears that it is vulnerable to misuse by school cheats or for disinformation.

    ‘ Not inexpensive’

    The dizzying success is in part because of OpenAI’s smart marketing strategy in, which it made its research accessible to non-experts, said AI expert Robb Wilson, creator of OneReach.ai, a software company.

    According to Wilson, having this tech available to technologists was one thing. Presenting it in a chat user interface and enabling non-developers to begin playing with it ignited a conversation.

    Founded in late 2015, OpenAI is conducted by Sam Altman, a 37-year-old entrepreneur and also the former president of startup incubator Y Combinator.

    The company has counted on the economic support of distinguished contributors from the start, including LinkedIn co-founder Reid Hoffman, investor Peter Thiel, and Musk.

    The multi-billionaire served on OpenAI’s board until 2018; however left to focus on Tesla, the electric automobile company.

    The startup also relies on a group of computer researchers and scientists led by Ilya Sutskever, a former Google executive that specializes in machine learning.

    OpenAI, which didn’t respond to AFP’s inquiries, had around 200 employees by 2021, according to a query made directly on ChatGPT.

    Meanwhile, despite the excitement generated by ChatGPT, the company has yet to discover a path to financial independence.

    Founded as a nonprofit, the startup turned into a “capped for-profit” company in 2019 to draw more investors. This week, co-founder Greg Brockman stated that a paid variation of ChatGPT was in the works.

    The quest for funding seems necessary for a company with outrageous expenses.

    In a Twitter exchange with Musk in early December, Altman acknowledged that each discussion on ChatGPT costs OpenAI many US cents.

    Accordingly to estimates by Tom Goldstein, an associate professor in the University of Maryland’s computer science division, the company is shelling out $100,000 a day for its bot or regarding $3 million a month.

    Partnering with Microsoft, which offers the startup with its remote computing services, might cut costs; however, “either way, it’s not cheap,” Goldstein stated.

    “Some state it’s wasteful to pour these kinds of resources … into a demo,” he included.

    Read the original article on SCIENCE ALERT.

    Read more: Monkeys– Not Humans– Made Old Sets of Stone Tools in Brazil, Study Discovers