How Could We Accelerate Global Bible Translation If We Started With a Blank Slate?

What if we started from the beginning, translating our first-ever Bible? How would we design a language model built specifically for low-resource contexts?

These are the questions the ETEN Innovation Lab is exploring in partnership with technology developers. Together, we are testing new translation processes assisted by cutting-edge artificial intelligence, intentionally built for the challenges and opportunities of under-resourced contexts.

Around the world, Bible translation efforts have often required tradeoffs. Partners and churches must navigate a balance between speed, affordability, and quality. Traditionally, only two of these objectives can be achieved at any one time.

A new approach pursues all three of those factors at the same time. By combining AI-assisted tools, modular software design, and community-led implementation, we are working towards scalable translation models. Tools like LangQuest and Codex are part of this vision, helping accelerate progress toward reaching the All Access Goals by 2033.

---

What If We Made New Designs From Scratch?

The blank slate framework emphasizes first identifying the greatest needs and bottlenecks in Bible translation. Then, we assess how our new technological capabilities can be leveraged to relieve these bottlenecks. To create new systems and processes for translating Scripture, the team returned to the basics to identify where inefficiencies actually lie. What they found is that the primary obstacles are often inherent in the structure and expectations of the translation process itself, rather than in the technologies available.

To design a model that is high-quality, fast, affordable, and, most importantly, scalable and replicable, the team posed a series of critical questions during the discovery phase:

• What happens if we clear the slate entirely and begin again?

• Can AI tools be purpose-built to meet the specific demands of challenging contexts and low-resource languages?

• Can we move beyond incremental gains and design for radical acceleration from the outset?

---

Identifying Bottlenecks

The team set out to partner with several cohorts of believers worldwide who expressed a desire for a Bible but were not candidates for traditionally managed translation projects. By attempting to discover the needs and capacities of these cohorts, the team uncovered the following two major bottlenecks in Bible translation:

1. Data scarcity

Many unreached communities lack existing Scripture or language data, which significantly limits the effectiveness of AI training workflows. Machine translation systems require examples of the target language to function, so any meaningful use of technology must begin by solving this problem. Accelerating Bible translation in low-resource contexts depends on our ability to gather that data quickly and effectively.

2. Adoption friction 

Even when tools are available, adoption rates among local translators and churches often remain low. Without a clear and replicable path to scalable adoption, these tools will fail to reach their full capabilities.

---

Overcoming Bottlenecks

To address these challenges, the team identified three key solutions:

1. Develop a rapid translation-data-gathering tool: Because low-resource languages are often spoken in regions with limited infrastructure, such as areas that are too hot, wet, or remote for laptops, this tool must be optimized for mobile phones and function reliably during extended periods without internet access.

2. Work with decentralized partners to scale adoption: Success in gathering data will only scale through on-the-ground networks and word-of-mouth sharing. The team is collaborating with disciple-making movements to reach as many languages and communities as possible in the coming years.

3. Create user-friendly AI-powered toolchains and platforms for leveraging this new data: Once sufficient data is collected for a new language, it must be used effectively. AI-powered tools can then apply this data to support and accelerate Bible translation workflows.

---

LangQuest

LangQuest is one of the early tools to emerge from applying a blank slate mindset to addressing the three identified challenges of Bible translation, which is currently in a development and experimentation phase. It is a mobile-first application designed for rapid translation-data collection by utilizing offline, multimodal translation workflows that incorporate text, images, and audio. Built specifically for low-resource environments, LangQuest collects training data by enabling community-led translation efforts where traditional infrastructure is limited or unavailable.

The app is optimized for long-term offline use. All data is stored locally on Android devices, allowing translators to work for weeks or even months without internet access. When reconnected, they can sync their fully intact projects, ensuring consistent progress regardless of connectivity.

LangQuest operates on the premise that through easy-to-use translation and validation tasks, the data collected can be used to train AI models in new languages. The experience follows a game-like format where teams translate, collect “gems”, receive feedback, and apply real-time learning as they work through “quests”. Users can listen to their translations for review, with validation by individuals other than the original translator to ensure high-quality, accurate translation data.

LangQuest aims to collect real-world training data by creating scalable, decentralized translation workflows. It supports translation “quests” of both biblical content and other local materials, as chosen by the user. The process is straightforward: source text, human translation, and validation. All data resides in the public domain, which means it is available for exporting, repurposing, and even training large-scale AI models from Meta or Google.

With enough data, the LangQuest team envisions the possibility of generating a full Bible draft zero in just hours at a cost of only a few dollars. This represents a significant shift, especially when compared to the timelines that have historically defined Bible translation efforts.

---

Codex

Codex is an AI-assisted desktop-based editor that complements LangQuest and is currently in a development and experimentation phase. It supports local translation workflows by adding structure, searchability, and flexibility across a range of translation applications, from Bible content to video subtitles.

In practice, Codex is used to translate materials such as subtitles from The Chosen into target languages. As translators work through the subtitles, the underlying AI in Codex learns the new language—essentially training the system in real time. Once trained on a dataset like The Chosen, Codex can provide translation suggestions for Bible passages in the same language, sometimes at the level of a verse or even a chapter. As each portion is edited, Codex immediately incorporates those changes, refining its future suggestions with continuous, real-time learning.

Codex also includes built-in audio recording, which helps translators detect and resolve issues in their drafts. Its flexible design accommodates a wide range of translation styles. Because the software works offline, translators can continue their work without interruption, even in environments with limited connectivity. This ensures faster project timelines and more consistent progress

Together, Codex and LangQuest reflect the blank slate approach to Bible translation. They combine the strengths of artificial intelligence with community-first design to enable scalable, locally led translation efforts.

---

Scalability On a Budget

This experimental approach is demonstrating what affordability can look like at scale. So far, it has supported a deployment across 50 languages at an average cost of just $6.86 per chapter. As adoption increases, these per-user and per-language costs are expected to decrease further. While some projects will still require additional support infrastructure, the technology remains accessible and cost-effective. This opens the door for greater reach, helping more communities to access the Bible in their heart language.

The long-term goal extends beyond producing faster or more accurate drafts. At the heart of the Innovation Lab and its partners is a shared vision: to equip local churches, translators, and communities with tools that are free to use, easy to adapt, and designed for long-term sustainability. This approach allows communities to take full ownership of the translation process, preserve their cultural and linguistic legacy, and invite broader participation in this moment of opportunity.

---

Next Steps

LangQuest and Codex were both developed through the blank slate framework. Early field testing indicates that each tool is meeting real needs in local communities. The development team continues to refine the approach, strengthen partnerships, and gather input from translators working in real-world conditions.

• To learn more about LangQuest, access langquest.org.

• To learn more about Codex, access codexeditor.app.

• For updates, insights, or opportunities to collaborate, visit lab.eten.bible or reach out to us at lab@eten.bible.