Connect with us

Technology

Speechify’s Simba 3.2 Ranks #1 on Independent Artificial Analysis TTS Leaderboard: World’s Best Real-Time Voice Model, Above ElevenLabs, OpenAI, Google DeepMind & Others

Published

on

Simba 3.2, the flagship streaming text-to-speech model from Speechify’s developer platform, is ranked first on the independent Artificial Analysis TTS leaderboard, above flagship models from ElevenLabs, Cartesia, OpenAI, and Google DeepMind, and joint second on Voice Arena’s blind listener rankings, making it the highest-rated real-time text-to-speech model and the least expensive in the leaderboard’s top ten, at $10 per million characters, dropping to $6 per million at scale.

NEW YORK, July 7, 2026 /PRNewswire-PRWeb/ — Speechify, the voice AI company behind the SpeechifyAI developer platform, today announced that its flagship text-to-speech model, Simba 3.2, has reached first place on the Artificial Analysis text-to-speech leaderboard, ranking above flagship models from ElevenLabs, Cartesia, OpenAI, and Google DeepMind, and joint second place on Voice Arena, two independent benchmarks that evaluate TTS models across the industry.

Simba 3.2 is ranked first on the Artificial Analysis text-to-speech leaderboard and joint second on Voice Arena’s blind listener rankings, making it the highest-rated real-time text-to-speech model available at its $6 per million character starting price.

Neither benchmark is operated by Speechify, and neither uses self-reported numbers. Both leaderboards update as new models ship, and both placed Simba 3.2 at or near the top of the field this week.

What the Rankings Show

The Artificial Analysis leaderboard (https://artificialanalysis.ai/text-to-speech/leaderboard) runs a consistent, objective evaluation across commercially available text-to-speech models and is widely tracked across the voice AI industry. Simba 3.2 currently holds first place, ahead of every evaluated model from ElevenLabs, Cartesia, OpenAI, Google DeepMind, and dozens of other commercial voice AI providers.

Voice Arena (https://voicearena.com/) ranks models through blind listener voting. Simba 3.2 currently sits joint second. Of the models ranked at or above it, one does not operate in real time, and the model tied with it is priced at roughly seven times Simba 3.2’s rate. Among real-time models that teams can deploy in production today, Simba 3.2 is the highest-rated option at its price point: $10 per one million characters on its entry tier, and $6 per one million characters on its Scale tier.

How Voice Arena Evaluates Text to Speech

Voice Arena uses the same evaluation approach that Chatbot Arena established as the standard for ranking large language models. Panels of native speakers hear two audio clips generated from identical text, without knowing which model produced which, and vote for the clip that sounds more natural. Votes are aggregated into an Elo rating for each model.

The methodology (https://voicearena.com/tts-methodology) is stricter than most public TTS evaluations. It covers six languages, uses a balanced voice slate per model rather than each vendor’s best-sounding default, and evaluates sentences written for the contexts where text to speech actually ships. The methodology was developed with advice from Prof. Shinji Watanabe of Carnegie Mellon University.

There is no self-reported mean opinion score, no vendor-selected sample, and no internal evaluation. The rankings reflect objective measurement and blind human listening.

Why the Result Matters for Real-Time Applications

For teams building voice agents, live readers, phone systems, and other applications where the voice responds in real time, a model that cannot stream is not a candidate regardless of how it sounds. Removing non-real-time models from the Voice Arena rankings leaves Simba 3.2 as the top-rated voice a team can put in production today at its price.

Best-sounding real-time and lowest-cost do not usually describe the same model. At $10 per one million characters on its entry tier, the price Artificial Analysis lists on the leaderboard, Simba 3.2 is the least expensive model in the leaderboard’s top ten. On the Scale tier the rate drops to $6 per one million characters, pricing built for production volume rather than demos. Full pricing is available at https://speechify.ai/pricing.

“This is the underdog story for API providers,” said Luke Oliff, Head of Developer Relations at Speechify. “We spent years making our models run efficiently because our consumer business demanded it, tens of millions of listeners, with some of the best voices on the planet. That work is why we can now put the best-rated model in the world on our API at about as cheap as it comes. Most labs built for the benchmark and priced for the enterprise. We built for listeners and priced for production.”

“As an undergrad at Stanford, CS229 with Andrew Ng sparked my interest in ML. My course project was fine-tuning Tacotron 2. My team’s new model at Speechify just hit SOTA, five years later. It’s pretty surreal looking back,” said Tyler Weitzman, Co-Founder, President and Head of AI at Speechify, in a post on X.

What Simba 3.2 Offers Developers

Simba 3.2 is a streaming-native model with lower time-to-first-byte than previous generations, fine-grained emotional control, and SSML prosody support. Emotion is modeled at the prosody level, covering the rhythmic and tonal patterns that convey feeling rather than only speed and pitch. The model supports instant voice cloning from short reference clips and native-quality speech across more than 30 locales, with mixed-language input handled automatically.

The model is available through the SpeechifyAI platform via a REST API and first-party TypeScript and Python SDKs, documented at https://docs.speechify.ai. This is the same API that powers Speechify’s consumer products, serving tens of millions of users.

Migration Support for Teams on Other Providers

For teams currently shipping on other text-to-speech providers, the gap between a better-sounding, lower-cost model and an actual switch involves real engineering work: voices to re-map, SSML to port, and latency to re-verify under production load. Speechify provides tooling and support to close that gap.

The Speechify Cookbook (https://github.com/SpeechifyInc/speechify-api-cookbook) is an open-source repository of runnable migration recipes, with SDK and native REST versions side by side so engineering teams can see exactly what goes over the wire. For customers with production volume, Speechify’s forward-deployed engineers (https://speechify.ai/forward-deployed-engineers) embed with the customer’s team and handle voice mapping, prosody parity, load testing, and cutover until quality holds at scale.

Availability

Simba 3.2 is available now through the SpeechifyAI platform. Developers can create a free API key at https://platform.speechify.ai and compare Simba 3.2 against any other model on their own text. Current rankings are published at https://artificialanalysis.ai/text-to-speech/leaderboard and https://voicearena.com, where anyone can participate in blind listener voting.

About Speechify

Speechify is a leading productivity platform serving more than 60 million users worldwide. Its product ecosystem includes Text to Speech, Voice Typing Dictation, AI Podcasts, the Voice AI Assistant, and Speechify Work, an offering that lets professionals delegate complex knowledge work to a team of AI agents. In 2025, Speechify received the Apple Design Award at WWDC, recognized as a critical resource for accessibility and productivity. The SpeechifyAI developer platform provides the same voice technology as an API, including streaming text to speech, instant voice cloning, emotional control, and support for more than 30 languages. Learn more at https://speechify.ai.

Media Contact

Rohan P, Speechify, 1 (747) 302-4454, rohan@speechify.com, speechify.ai

View original content to download multimedia:https://www.prweb.com/releases/speechifys-simba-3-2-ranks-1-on-independent-artificial-analysis-tts-leaderboard-worlds-best-real-time-voice-model-above-elevenlabs-openai-google-deepmind–others-302819731.html

SOURCE Speechify

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Technology

Debevoise Expands Fund Finance Practice with the Return of Zahra Sowder as Partner

Published

on

By

NEW YORK, July 14, 2026 /PRNewswire/ — Debevoise & Plimpton LLP announced today that Zahra Sowder has joined the firm as a partner in its Fund Finance practice in New York.

Ms. Sowder advises private funds and financial institutions on a broad range of complex fund financing transactions, including subscription lines and NAV facilities, as well as bespoke financing solutions for LP and GP stakes. She also has extensive experience advising corporate borrowers on asset-based financings.

Presiding Partner Peter Furci said, “We continue to invest in our private capital platform to meet growing client demand for sophisticated financing solutions across the fund lifecycle. Zahra’s deep experience in this area, together with her longstanding connection to Debevoise and many of our clients, makes her an exceptional addition to the firm.”

Co-Chair of the Fund Finance Group Ramya Tiller said, “Private capital clients need sophisticated financing advice to support liquidity, fundraising and investment activity. Zahra has deep experience across the products our clients are using most actively, including NAV facilities across all strategies and other complex fund financing solutions. Having previously practiced at Debevoise, Zahra can contribute immediately to our growing practice.”

Ms. Sowder said, “Debevoise played an important role earlier in my career and is ideally suited for this next stage of my practice, especially given its specialized and growing fund finance practice. I look forward to working with colleagues and clients I know well and contributing to the continued growth of the practice.”

Ms. Sowder joins Debevoise from the New York office of another international law firm and previously practiced at Debevoise for more than 15 years as an associate and counsel. She received her J.D. from UCLA School of Law in 2008 and her B.A. from Vassar College in 2002.

View original content:https://www.prnewswire.com/news-releases/debevoise-expands-fund-finance-practice-with-the-return-of-zahra-sowder-as-partner-302825501.html

SOURCE Debevoise & Plimpton LLP

Continue Reading

Technology

As AI Agents Scale, Enterprises Demand Execution Control — Devenex Takes Control

Published

on

By

LAS VEGAS, July 14, 2026 /PRNewswire/ — • Inside the world’s largest organisations, AI agents are no longer experimental. They are executing in production — modifying financial records, triggering payments, approving workflows, and acting with the full operational authority of the enterprises that deployed them. By every credible estimate, the volume and consequence of these actions will increase by an order of magnitude within 24 months.

Yet no infrastructure layer exists to govern what these agents actually do. No systematic policy enforcement before action. No immutable audit trail after. No deterministic control between an agent’s intent and its real-world consequence. The most consequential technology shift in enterprise history is unfolding without an accountability layer.

Today, that changes. Devenex — the Execution Control Plane for AI agents — launches at Google Cloud Next 2026, introducing enterprise-grade governance infrastructure that sits between agent intent and real-world execution. Every action is policy-evaluated, explicitly authorised, identity-bound, and recorded as audit-grade evidence — before it takes effect, giving new life to enterprise AI initiatives.

Devenex is not a monitoring tool. It is not a workflow engine. It is the control plane that ensures no AI-initiated action executes without governance — giving CIOs, CTOs, CROs and enterprise security leaders the confidence to deploy agents at scale without sacrificing accountability.

THE STRUCTURAL GAP

Analyst consensus has converged on a single conclusion: enterprise AI execution capability is outpacing the controls designed to keep it accountable.

Gartner projects that by 2028, a third of enterprise software will incorporate agentic AI — up from less than one percent in 2024. McKinsey identifies governance and risk as the primary barriers to scaling enterprise AI, ahead of model quality or talent. Deloitte’s 2026 enterprise survey finds that 80 percent of leaders piloting AI agents cite security and compliance as the leading obstacle — up from 68 percent a year earlier.

The pattern is consistent across every credible source: enterprises are deploying agents faster than they can govern them. An AI agent that modifies a customer record, approves a discount, or initiates a wire transfer without a governing policy layer is not automation. It is an unmanaged compliance event, a latent security exposure, and a board-level liability.

Devenex addresses this gap at the infrastructure level — not as a feature bolted onto existing tools, but as a purpose-built control plane designed for governed enterprise execution from the ground up.

THE EXECUTION CONTROL PLANE

Devenex operates as the authorisation and governance layer between enterprise decision-making — whether initiated by humans, AI agents, or automated systems — and the execution of actions across enterprise systems of record.

Every action processed through Devenex produces four structured artifacts:

Intent Record. 

Execution Plan. 

Governed Execution. 

Execution Evidence. 

This model ensures that enterprises maintain full traceability from intent to outcome — satisfying EU AI Act, SOC 2, ISO 42001, and sector-specific regulatory requirements.

ENTERPRISE CAPABILITIES

Pre-Execution Policy Enforcement.  Every agent action is evaluated against organizational policy before execution. Nothing executes unchecked. Enterprises gain control over what agents are permitted to do — reducing compliance risk and eliminating ungoverned action.

Dynamic Human-in-the-Loop Governance.  High-consequence actions are routed to designated reviewers without halting low-risk automation. Approval workflows are configurable at the agent, action, or policy level — ensuring human oversight where it matters without creating operational bottlenecks.

Immutable Audit Infrastructure.  Every governed execution is recorded to an append-only ledger. Enterprises gain continuous compliance evidence without manual reporting.

Unified Observability Across Agents.  Live and retrospective visibility into agent activity, policy adherence, and anomaly patterns. Security, compliance, and operations teams share a single authoritative view of enterprise execution.

BUILT FOR ENTERPRISE REALITY

Devenex delivers flexible deployment models—SaaS, hybrid, and self-deployed—enabling organizations to adopt at their own pace.

Engineered to be framework-agnostic and cloud-native, Devenex integrates seamlessly across diverse enterprise environments. The platform does not replace existing systems of record, integration platforms, or identity providers. It governs execution across them.

LEADERSHIP

“For four decades, Abacus has earned the trust of enterprises navigating their most consequential technology transitions. Devenex represents the next chapter — purpose-built infrastructure for a world where AI agents execute with the authority of the organisations that deploy them. Governance at this layer is not optional. It is a precondition for enterprise AI at scale.”

— Aly Kuly Khan
Co-Founder & Chairman, Devenex

“For four decades, we’ve built the layers enterprises run on — systems of record, integration, workflow automation, API and iPaaS governance. Each wave solved the problem the previous one created. Today, AI agents are executing actions on architecture that was never designed to govern them. This isn’t an AI problem. It’s an architectural gap — and it’s the one our experience has prepared us to solve. Enterprises cannot answer four questions about any agentic action: who authorized it, what policy governed it, why it executed as it did, and whether they can prove it after the fact. In regulated environments, these aren’t edge cases — they’re the baseline. Devenex is the execution control plane that answers all four, by design, at execution time.”

— Shoaib A. Khan
Co-Founder & CEO, Devenex

AVAILABILITY & ENGAGEMENT

Devenex is available immediately, with deployment support from Abacus teams across the globe.

Enterprise pilot programmes are open to qualified organisations seeking to bring governed execution to production AI workloads.

Learn more at www.devenex.com

THE ROAD AHEAD

Enterprise AI is entering its execution era. The question is no longer whether agents can act — it is whether they can act accountably. The organisations that solve this first will scale AI faster, operate with greater confidence, and carry less risk than those that treat governance as a downstream problem.

Devenex exists to make governed execution the default operating model of the enterprise. Not as an aspiration. As infrastructure.

About DevenEx

Devenex is the Execution Control Plane for AI agents — enterprise infrastructure that governs every action across systems with policy enforcement, explicit authorisation, and audit-grade evidence. Built for the agentic era, Devenex sits between intent and execution so that no enterprise action moves forward ungoverned. Devenex is built by the team behind Abacus, bringing four decades of enterprise trust to AI execution governance.

About Abacus

Abacus is a global professional services leader in technology, outsourcing, and people solutions. With nearly 40 years of experience, 5,000+ professionals across four continents, and 1,500+ enterprise clients, Abacus designs bespoke solutions that enable organisations to create the future of business and embrace change for sustainable growth.

Media Contact

Devenex
www.devenex.com
Shoaib Khan
shoaib.khan@abacuscambridge.com
+1 (347) 701-4221

View original content to download multimedia:https://www.prnewswire.co.uk/news-releases/as-ai-agents-scale-enterprises-demand-execution-control–devenex-takes-control-302825507.html

Continue Reading

Technology

CASIO Makes Classroom-Ready Calculator Free for Everyone with Launch of fx-300ES PLUS Online Emulator

Published

on

By

Casio’s fx-300ES PLUS Online Emulator is the First Web-Accessible Emulator of the Iconic Scientific Calculator, Giving Students, Parents and Teachers Free, Instant Access

JERSEY CITY, N.J., July 14, 2026 /PRNewswire/ — Casio America, Inc. today announced a free online emulator for its fx-300ES PLUS scientific calculator, available now at ClassPad.net. The first web-accessible emulator of its kind for the fx-300ES series gives students and educators instant access to a fully functional digital calculator—at no cost—from any internet-connected device.

“The 300ES PLUS has been a staple in math and science classrooms for years because it’s reliable and easy to use,” said Marko Wityk, Senior General Manager of EdTech at Casio America. “This emulator takes that same experience and makes it available to any student. Whether they’re finishing homework or teachers are walking a class through a problem, this tool removes any barriers standing in the way of learning.”

Designed for Classrooms

The emulator requires only a free login through Google, Google Classroom, Microsoft, or Clever, platforms already in use at schools nationwide. Users gain access to a fully functional digital calculator replicating the fx-300ES PLUS, along with other scientific and graphing calculator options from Casio.

Key features include:

Projection mode – The emulator screen can be enlarged for display on a projector or interactive whiteboard, letting teachers walk the entire class through problems step-by-step.Key log – Displays a running log of each key used in a problem, allowing students to easily follow along with the teacher.Natural Textbook Display – See math problems and fractions exactly as they would appear in textbooks, homework, or exams.Zero cost and instant access – No app download needed and no purchase or subscription required now or in the foreseeable future.Device-agnostic – Compatible with any web-enabled device, from school Chromebooks to home laptops to tablets and other mobile devices.Screenshot – Capture an image of the calculator display to compare results and functions. Screenshots are copied to the clipboard for use in other applications as well.

The fx-300ES PLUS emulator is available now, free of charge, at ClassPad.net. Users can log in with an existing Google, Google Classroom, Microsoft, or Clever account. No additional registration is required.

The physical calculator remains available at major retailers nationwide and online at casio.com, typically priced around $15. Education pricing is available to teachers and schools by contacting edtech@casio.com.

About Casio America, Inc.
Casio America, Inc. is a subsidiary of Casio Computer Co., Ltd., one of the world’s leading manufacturers of consumer electronics and business equipment solutions. Casio’s EdTech division is dedicated to empowering students, educators, and parents with tools that make learning more accessible and effective. Casio products are sold through authorized Casio dealers. For more information, visit casio.com.

View original content to download multimedia:https://www.prnewswire.com/news-releases/casio-makes-classroom-ready-calculator-free-for-everyone-with-launch-of-fx-300es-plus-online-emulator-302825523.html

SOURCE Casio America, Inc.

Continue Reading

Trending