Speaking Assessment with Zero Exam Fees and Grading Time
Key facts
- Speaking Assessment with Zero Exam Fees and Grading Time
- ELSA Japan announces the launch of its AI-powered Speaking Assessment feature for 'ELSA School' in summer 2026, enabling teachers to objectively evaluate students' English speaking skills in class without additional costs or grading time.
- Source: PR Times
- Date: June 18, 2026
Direct answer
ELSA Japan announces the launch of its AI-powered Speaking Assessment feature for 'ELSA School' in summer 2026, enabling teachers to objectively evaluate students' English speaking skills in class without additional costs or grading time.
- Citation
- Speaking Assessment with Zero Exam Fees and Grading Time (June 18, 2026), PR Times
- Source
- PR Times
- Date
- June 18, 2026
ELSA Japan announces the launch of its AI-powered Speaking Assessment feature for 'ELSA School' in summer 2026, enabling teachers to objectively evaluate students' English speaking skills in class without additional costs or grading time.
📋 Article Processing Timeline
- 📰 Published: June 18, 2026 at 00:43
- 🔍 Collected: June 17, 2026 at 15:47
- 🤖 AI Analyzed: June 18, 2026 at 15:33 (23h 45m after Collected)
ELSA Japan LLC (Shibuya, Tokyo; Representative: Daisuke Imao), operator of the AI English learning service 'ELSA School,' announces the launch of its 'AI Speaking Assessment' feature in summer 2026. This feature enables teachers to objectively measure students' English speaking abilities within regular classroom instruction.
With this feature, homeroom and English teachers can continuously monitor each student's speaking proficiency without relying on external certification exams, seamlessly integrating assessment into daily lessons.
Background: The 'Assessment Gap' Facing Educators
Starting in the 2026 academic year, the National Assessment of Academic Ability and Learning Status will include a CBT-based 'speaking' component for junior high school English, highlighting the growing institutional importance of speaking instruction. However, a key challenge remains: the lack of practical tools for ongoing, in-class effectiveness measurement.*2
While external English proficiency tests offer reliable benchmarks, they come with significant costs—including exam fees, implementation expenses, and preparation time. Regularly administering these tests to all students is often impractical for many schools. On the other hand, assessments using generic speech recognition or generative AI often fail to clearly identify issues related to pronunciation, fluency, and clarity.
The 'Speaking Assessment' feature in ELSA School was designed to fill this gap. Leveraging ELSA’s proprietary speech recognition engine and a global dataset of over 90 million non-native English utterances from 190 countries, the system analyzes student speech at the phoneme level, instantly identifying 'where and how much' pronunciation deviates. This allows teachers to save time on grading and focus on improving instruction.
AI Assessment Feature Overview
1. Adaptive Question Design
Difficulty dynamically adjusts based on student responses, enabling accurate assessment for all students—even in classrooms with wide proficiency ranges—within a short time.
2. CEFR-Aligned Accurate Feedback
Instantly delivers scores aligned with the CEFR (Pre-A1 to C2), the international standard for language proficiency. Provides detailed feedback across five dimensions—pronunciation, intonation, fluency, vocabulary, and grammar—to support student speaking improvement.
3. Zero Additional Costs, Unlimited Attempts
Schools using ELSA School can administer the assessment an unlimited number of times at no extra cost—no exam fees or additional licensing required. This facilitates formative assessments each term, pre- and post-instruction comparisons, and personalized learning cycles.
The feature is currently being piloted in select schools and will be rolled out to all ELSA School users starting summer 2026.
Technical Foundation for Accurate Evaluation: In-House Engine Ensures Consistent Quality
The market for English speaking assessment tools has seen a rise in services built using generic speech recognition APIs and generative AI. While this approach reduces development costs, it heavily depends on external engines’ specifications, performance, and update policies, making it difficult to guarantee the consistent, fine-grained pronunciation evaluation required in educational settings.
ELSA has maintained and operated its own proprietary speech recognition engine—specialized in non-native English pronunciation evaluation—since its founding, along with a globally collected English speech dataset. The phoneme-level analysis accuracy is achieved through this in-house infrastructure, enabling continuous quality improvements without reliance on external engines.
This technical foundation is supported by investments led by Gradient Ventures, Google’s AI-focused corporate venture capital firm, providing a strong financial base for sustained R&D investment—a key differentiator from competitors.
AI tools adopted in schools must not only work today but remain reliable and stable for years to come. ELSA’s in-house technology and financial backing serve as the foundation for this long-term trust.
World-Class Security Infrastructure: Ensuring Safe Use for Schools and Municipalities
ELSA holds globally recognized certifications for personal data handling. It meets standard security requirements sought by municipalities and school corporations in procurement—such as the Ministry of Education’s AI utilization initiatives—eliminating concerns about post-implementation data management.
The company holds four major certifications: GDPR (EU General Data Protection Regulation), CCPA (California Consumer Privacy Act), SOC 2 (third-party security audit), and ISO 27001:2022 (international information security standard). Data transfer, storage, and access controls are subject to ongoing audits by international third-party organizations.
■ Stakeholder Comments
Miho Toda, Senior Officer and Instructional Supervisor, Kyotango City Board of Education
'We’ve been looking for an affordable and fast way to regularly measure CEFR levels in class. ELSA’s new assessment feature perfectly meets this need. We expect it to provide more accurate insights into students’ CEFR levels and enhance daily instruction.'
Shun Yanagida, Chief Instructional Supervisor, Shibuya City Board of Education
'Being able to more accurately and objectively understand students’ learning progress enables more adaptive teaching. Introducing this feature—which reduces teacher assessment burden while providing personalized feedback—marks a major advancement for English education in Shibuya City.'
Daisuke Imao, Representative, ELSA Japan LLC
'With speaking now included in the national academic assessment, Japan’s English education is at a turning point. We aim to address the challenge of 'assessment difficulty' in schools by offering an objective, accessible solution powered by AI. Through this feature, we hope to empower students nationwide to recognize their growth and confidently engage with the world.'
■ Company Overview
Company Name: ELSA Japan LLC
Representative: Daisuke Imao
Head Office: 16F, Link Square Shinjuku, 5-27-5 Sendagaya, Shibuya-ku, Tokyo
Business: Development and provision of AI-powered English conversation learning apps and English proficiency support services
Website: https://jp.elsaspeak.com/
________________
*1 Source: Internal data: Correlation coefficient of r=0.897 between EPS and evaluations by IELTS certified examiners on 52 speech samples. 94.23% of predictions within ±1.0 IELTS band.
*2 Source: Ministry of Education, Culture, Sports, Science and Technology, 'National Assessment of Academic Ability and Learning Status, Reiwa 8' (to be conducted April 24–May 29, 2026).
With this feature, homeroom and English teachers can continuously monitor each student's speaking proficiency without relying on external certification exams, seamlessly integrating assessment into daily lessons.
Background: The 'Assessment Gap' Facing Educators
Starting in the 2026 academic year, the National Assessment of Academic Ability and Learning Status will include a CBT-based 'speaking' component for junior high school English, highlighting the growing institutional importance of speaking instruction. However, a key challenge remains: the lack of practical tools for ongoing, in-class effectiveness measurement.*2
While external English proficiency tests offer reliable benchmarks, they come with significant costs—including exam fees, implementation expenses, and preparation time. Regularly administering these tests to all students is often impractical for many schools. On the other hand, assessments using generic speech recognition or generative AI often fail to clearly identify issues related to pronunciation, fluency, and clarity.
The 'Speaking Assessment' feature in ELSA School was designed to fill this gap. Leveraging ELSA’s proprietary speech recognition engine and a global dataset of over 90 million non-native English utterances from 190 countries, the system analyzes student speech at the phoneme level, instantly identifying 'where and how much' pronunciation deviates. This allows teachers to save time on grading and focus on improving instruction.
AI Assessment Feature Overview
1. Adaptive Question Design
Difficulty dynamically adjusts based on student responses, enabling accurate assessment for all students—even in classrooms with wide proficiency ranges—within a short time.
2. CEFR-Aligned Accurate Feedback
Instantly delivers scores aligned with the CEFR (Pre-A1 to C2), the international standard for language proficiency. Provides detailed feedback across five dimensions—pronunciation, intonation, fluency, vocabulary, and grammar—to support student speaking improvement.
3. Zero Additional Costs, Unlimited Attempts
Schools using ELSA School can administer the assessment an unlimited number of times at no extra cost—no exam fees or additional licensing required. This facilitates formative assessments each term, pre- and post-instruction comparisons, and personalized learning cycles.
The feature is currently being piloted in select schools and will be rolled out to all ELSA School users starting summer 2026.
Technical Foundation for Accurate Evaluation: In-House Engine Ensures Consistent Quality
The market for English speaking assessment tools has seen a rise in services built using generic speech recognition APIs and generative AI. While this approach reduces development costs, it heavily depends on external engines’ specifications, performance, and update policies, making it difficult to guarantee the consistent, fine-grained pronunciation evaluation required in educational settings.
ELSA has maintained and operated its own proprietary speech recognition engine—specialized in non-native English pronunciation evaluation—since its founding, along with a globally collected English speech dataset. The phoneme-level analysis accuracy is achieved through this in-house infrastructure, enabling continuous quality improvements without reliance on external engines.
This technical foundation is supported by investments led by Gradient Ventures, Google’s AI-focused corporate venture capital firm, providing a strong financial base for sustained R&D investment—a key differentiator from competitors.
AI tools adopted in schools must not only work today but remain reliable and stable for years to come. ELSA’s in-house technology and financial backing serve as the foundation for this long-term trust.
World-Class Security Infrastructure: Ensuring Safe Use for Schools and Municipalities
ELSA holds globally recognized certifications for personal data handling. It meets standard security requirements sought by municipalities and school corporations in procurement—such as the Ministry of Education’s AI utilization initiatives—eliminating concerns about post-implementation data management.
The company holds four major certifications: GDPR (EU General Data Protection Regulation), CCPA (California Consumer Privacy Act), SOC 2 (third-party security audit), and ISO 27001:2022 (international information security standard). Data transfer, storage, and access controls are subject to ongoing audits by international third-party organizations.
■ Stakeholder Comments
Miho Toda, Senior Officer and Instructional Supervisor, Kyotango City Board of Education
'We’ve been looking for an affordable and fast way to regularly measure CEFR levels in class. ELSA’s new assessment feature perfectly meets this need. We expect it to provide more accurate insights into students’ CEFR levels and enhance daily instruction.'
Shun Yanagida, Chief Instructional Supervisor, Shibuya City Board of Education
'Being able to more accurately and objectively understand students’ learning progress enables more adaptive teaching. Introducing this feature—which reduces teacher assessment burden while providing personalized feedback—marks a major advancement for English education in Shibuya City.'
Daisuke Imao, Representative, ELSA Japan LLC
'With speaking now included in the national academic assessment, Japan’s English education is at a turning point. We aim to address the challenge of 'assessment difficulty' in schools by offering an objective, accessible solution powered by AI. Through this feature, we hope to empower students nationwide to recognize their growth and confidently engage with the world.'
■ Company Overview
Company Name: ELSA Japan LLC
Representative: Daisuke Imao
Head Office: 16F, Link Square Shinjuku, 5-27-5 Sendagaya, Shibuya-ku, Tokyo
Business: Development and provision of AI-powered English conversation learning apps and English proficiency support services
Website: https://jp.elsaspeak.com/
________________
*1 Source: Internal data: Correlation coefficient of r=0.897 between EPS and evaluations by IELTS certified examiners on 52 speech samples. 94.23% of predictions within ±1.0 IELTS band.
*2 Source: Ministry of Education, Culture, Sports, Science and Technology, 'National Assessment of Academic Ability and Learning Status, Reiwa 8' (to be conducted April 24–May 29, 2026).
FAQ
Who can use this assessment feature?
Middle and high schools nationwide using ELSA School, including boards of education. Teachers and students can use it in class.
Can assessment results be used for grading?
Yes, results serve as formative assessment data and can support official grading decisions.
Is internet access required?
Yes, stable internet is required for real-time cloud-based speech analysis.