FAQs

AskThreeBots lets you ask three separate AI models. It compares each answer. Then it summarises the similarities and differences. 

If there is an agreement between two or more models, a Consensus Summary will be shown. This summary will show how, and which models agree.

Then there will be no consensus. But you’ll still be given a summary showing how the answers differ between each AI model.

With AskThreeBots, you don’t have to rely on just one AI model. Neither do you have to spend time cross-checking between answers from different models. 

In time, AskThreeBots will also be able to detect biases within each AI model.

Yes. AskThreeBots has a large selection of the world's most commonly used Large Language Models (LLMs).

Just choose the one model that will analyse and formulate the three answers, and choose the three models that form your panel of bots.

Asking three bots means waiting for all three AI models to analyse and write their answer. Then a fourth model needs to analyse and look for a consensus. It’s not as fast as asking just one AI. But it’s definitely faster than reading three separate answers from three AI models.

Is there anything I can do to increase the speed of response?

Your response is only as fast as the slowest of your three bots. So by watching which bot is the slowest to respond, you can change the bot configuration accordingly.

• Software Engineers and CTOs

"Can I trust AI with my codebase?"

AI-generated code often includes hallucinated libraries or security protocols. Engineers use AskThreeBots as a logic-check layer. By running a three-bot consensus, they identify edge-case errors that a single model may overlook. If two bots suggest a solution but the third flags a memory leak, you've saved a week of debugging

• Medical Researchers & Health Tech Professionals

"How do I verify clinical summaries?"

Single AI models often struggle with contradictory study results. Researchers use our platform to perform consensus literature reviews. By querying three independent models, they ensure a summary isn't just a pattern hallucination, but more likely a factual reflection of the data.

• Investigative Journalists & Content Creators

"How do I protect my reputation from AI misinformation?"

Our media users employ AskThreeBots for triple-factor verification. Before publishing any high-stakes claim, they run it through our engine. If the bots diverge on a name, date, or event, it serves as a red flag to return to primary sources.

• Data Scientists & Business Analysts

"How do I spot bias in my market projections?"

AI models have "optimistic biases" based on their training sets. Analysts use AskThreeBots to identify statistical volatility. If three models provide three different interpretations of the same dataset, the analyst knows the data may be unreliable.

Every time you use our services, 1% of the transaction goes to Stripe Climate. This contribution supports a coalition of frontier companies working on permanent carbon removal, ensuring that our computational growth does not come at the Earth's expense.

AskThreeBots utilises a Sovereign Data Architecture. All primary user data is processed and stored in ISO 27001 Tier 3-certified data centres in Singapore.

Absolutely. We treat your data with a strict, privacy-first architectural approach:

• Advanced Encryption: All data transmissions are fully protected using industry-standard TLS 1.3 in transit and AES-256 encryption at rest.

• Privacy-First Routing: We utilize enterprise-grade API connections via OpenRouter, actively requesting Zero Data Retention (ZDR) pipelines wherever supported by the underlying frontier model providers.

• No Model Training: We explicitly do not use your prompts, code, or generated history to train our own or any third-party foundation models.

• Data Control: You maintain full authority over your logs. You can manage your account data or clear your operational chat history at any time directly through your dashboard settings.

The currency for Generative AI compute power is 'tokens'. The number of tokens depends on the complexity of the query. For example, asking “What’s the capital of Switzerland” will incur far fewer tokens than “Read my attached 100-page manuscript and provide a detailed synopsis”. 

AskThreeBots charges based on the collective compute used. Since each query runs a parallel panel of three elite models plus a consensus layer, we keep things simple by showing total cost as 'Consensus Credits'.  

You can purchase Consensus Credits by subscribing to a fixed monthly fee. And you can top-up with extra credits.

Credits purchased as part of the monthly subscription fee cannot be rolled over into the next month. However, credits bought as top-ups are valid for one full year after date of purchase, giving you flexibility for seasonal workloads.