Yozora Finance Launches "Momonga Search API" Beta, Enabling AI Agents to Access Corporate Disclosures and Economic News, Cross-Searching Over 3.2 Million Texts and Tables
Yozora Finance Co., Ltd. has launched the beta version of its "Momonga Search API," a REST API for developers designed to provide corporate disclosures and economic news in an AI-friendly format. It covers over 3,800 companies in major markets, including timely disclosures and annual securities reports, and cross-searches over 3.2 million texts and tables.
📋 Article Processing Timeline
- 📰 Published: May 11, 2026 at 19:00
- 🔍 Collected: May 11, 2026 at 10:31
- 🤖 AI Analyzed: May 12, 2026 at 05:44 (19h 12m after Collected)
Yozora Finance Co., Ltd. has commenced the beta provision of "Momonga Search API," a REST API for developers that enables AI to easily acquire corporate disclosures and economic news.
It covers over 3,800 companies in major markets, responding to all new timely disclosures and all annual securities reports. It also updates daily with materials only available on company websites, and news such as interest rates and geopolitics.
A 14-day free trial allows full access to all features without credit card registration.
Furthermore, by subscribing to a paid plan during the beta period (until May 24th), all plans will be available at a perpetual 50% discount as long as the subscription continues.
Service website: https://momongasearch.com
Background of Development
With the spread of AI/AI agents like ChatGPT and Claude Code, an environment is emerging where individuals can create their "personal investment research tools." However, many users encounter data preparation hurdles when continuously collecting corporate disclosures and news.
Cannot trust AI output
⇨ Web search mixes countless secondary and tertiary information (individual speculation, AI-generated articles not thoroughly vetted, etc.) with a single primary source.
Usage limits are quickly reached
⇨ Directly feeding large PDFs like financial statements and annual reports consumes context window and token budgets.
⇨ Low-quality data collected from the web wastes tokens and degrades response quality.
Data APIs handling timely corporate disclosures and economic news are limited.
⇨ Pricing is set for corporate use, making it unaffordable for individuals.
Momonga Search API is an API designed to eliminate such developer effort and waste.
Innovations in API design, born from 1 year of in-house AI utilization, optimized for AI use
Acquire financial statements in Markdown
Deep learning models detect layouts to capture headings, tables, charts, and formulas, converting this information into Markdown text.
Additionally, for pages containing visual information that cannot be fully expressed in text, the API pre-determines if it's necessary to acquire only that page as an image.
Search by keyword or meaning
Over 3.2 million texts and tables are pre-vectorized, allowing AI to quickly identify only the necessary sentences. Combining sparse vector search (strong in keyword matching) and dense vector search (strong in expression variation) enables meaning-based document exploration, which is difficult with grep search in AI agents like Claude Code.
Acquire only necessary sections from the table of contents
First, grasp the overview from the table of contents. You can then acquire only the required sections of the body. Since the character count per section is known beforehand, you can estimate the reading volume and cost before feeding it to an LLM.
Candidate search consumes no credits
Checking for new arrivals, acquiring titles, and tables of contents consume no credits. They can be called repeatedly within rate limits. Conveniently, identifying security codes also costs 0 credits. Even if AI specifies a company by abbreviation, old name, or notation variation, it can be identified using a dictionary of approximately 43,000 name-matching entries.
About the scope of provision
With Momonga Search API, you can check the availability of texts, images, original files, and reference URLs for each document in the API response.
For some documents, due to terms of service or rights from the data source, provision may be limited to metadata and reference URLs instead of texts or original files. Also, for some documents, the text may become available after a certain period has passed since publication. Please check the service website for details on the scope of support.
About Beta Provision
Momonga Search API is currently offered as a beta version. Users who subscribe to a paid plan during the beta period (until May 24th) will receive a perpetual 50% discount on their paid plan as long as their subscription continues.
The development team will continuously work on improving search quality, adding sample code, and expanding supported document and news types based on your feedback. Please send us your honest opinions.
[Operating Company & Contact]
Operating Company: Yozora Finance Co., Ltd.
Representative Director: Yusuke Tomita
Service website: https://momongasearch.com
Contact: contact@yozorafinance.com
It covers over 3,800 companies in major markets, responding to all new timely disclosures and all annual securities reports. It also updates daily with materials only available on company websites, and news such as interest rates and geopolitics.
A 14-day free trial allows full access to all features without credit card registration.
Furthermore, by subscribing to a paid plan during the beta period (until May 24th), all plans will be available at a perpetual 50% discount as long as the subscription continues.
Service website: https://momongasearch.com
Background of Development
With the spread of AI/AI agents like ChatGPT and Claude Code, an environment is emerging where individuals can create their "personal investment research tools." However, many users encounter data preparation hurdles when continuously collecting corporate disclosures and news.
Cannot trust AI output
⇨ Web search mixes countless secondary and tertiary information (individual speculation, AI-generated articles not thoroughly vetted, etc.) with a single primary source.
Usage limits are quickly reached
⇨ Directly feeding large PDFs like financial statements and annual reports consumes context window and token budgets.
⇨ Low-quality data collected from the web wastes tokens and degrades response quality.
Data APIs handling timely corporate disclosures and economic news are limited.
⇨ Pricing is set for corporate use, making it unaffordable for individuals.
Momonga Search API is an API designed to eliminate such developer effort and waste.
Innovations in API design, born from 1 year of in-house AI utilization, optimized for AI use
Acquire financial statements in Markdown
Deep learning models detect layouts to capture headings, tables, charts, and formulas, converting this information into Markdown text.
Additionally, for pages containing visual information that cannot be fully expressed in text, the API pre-determines if it's necessary to acquire only that page as an image.
Search by keyword or meaning
Over 3.2 million texts and tables are pre-vectorized, allowing AI to quickly identify only the necessary sentences. Combining sparse vector search (strong in keyword matching) and dense vector search (strong in expression variation) enables meaning-based document exploration, which is difficult with grep search in AI agents like Claude Code.
Acquire only necessary sections from the table of contents
First, grasp the overview from the table of contents. You can then acquire only the required sections of the body. Since the character count per section is known beforehand, you can estimate the reading volume and cost before feeding it to an LLM.
Candidate search consumes no credits
Checking for new arrivals, acquiring titles, and tables of contents consume no credits. They can be called repeatedly within rate limits. Conveniently, identifying security codes also costs 0 credits. Even if AI specifies a company by abbreviation, old name, or notation variation, it can be identified using a dictionary of approximately 43,000 name-matching entries.
About the scope of provision
With Momonga Search API, you can check the availability of texts, images, original files, and reference URLs for each document in the API response.
For some documents, due to terms of service or rights from the data source, provision may be limited to metadata and reference URLs instead of texts or original files. Also, for some documents, the text may become available after a certain period has passed since publication. Please check the service website for details on the scope of support.
About Beta Provision
Momonga Search API is currently offered as a beta version. Users who subscribe to a paid plan during the beta period (until May 24th) will receive a perpetual 50% discount on their paid plan as long as their subscription continues.
The development team will continuously work on improving search quality, adding sample code, and expanding supported document and news types based on your feedback. Please send us your honest opinions.
[Operating Company & Contact]
Operating Company: Yozora Finance Co., Ltd.
Representative Director: Yusuke Tomita
Service website: https://momongasearch.com
Contact: contact@yozorafinance.com