FREQUENTLY ASKED QUESTIONS

F68 AI is an AI Agent and AI Fintech platform.
At the AI Agent layer, F68 AI lets you use multiple models, build Agents with a System Prompt, Context and Tools, manage Context, and critique results with another AI.
At the AI Fintech layer, F68 AI is developing F68 Data → F68 Intelligence → F68 Trading, starting with F68 AI Stocks.

No.
Multi-model is one part of the product.
The core value of F68 AI lies in the Agent layer — where a model is combined with a System Prompt, Context, Tools and an operating layer to work by a specific method.

Yes. F68 calls exactly the model and configuration shown in the interface through the corresponding provider’s API, without simulating it or silently switching to a lower model.

Yes. You can choose the model and reasoning level that fit each task. Specialized AI Agents can use configurations preset by F68.

An AI Agent is not a new model.
It is a model placed in a working method designed in advance.
Users configure an Agent with:
System Prompt · Context · Tool
F68 AI runs the system layer needed for the Agent to work with Context, Tools and history in the same environment.

A personal AI Agent is set up by you with your own method and data. F68 Stocks is a system of specialized Agents developed by F68 AI for analyzing Vietnamese stocks.

One AI produces a result, and another AI critiques it independently to check assumptions, weaknesses and missing perspectives. It is not the same AI switching roles.

You top up Credit and pay by actual usage. The cost of each task is recorded after it completes and can be viewed in your account history.

Yes. Token rates follow the official API pricing of each vendor and are updated when the provider changes its prices.

F68 AI uses AI at enterprise scale and benefits from commercial policies and volume discounts, which lets it keep the price to users at the official API price.

Trial Credit is for eligible tasks on F68 AI, cannot be converted into cash, and is subject to the terms of each program.

Caching is a mechanism of the model/API provider that reuses Context that was already processed earlier when a new request meets the cache conditions.
A cache hit usually costs less input than a cache miss, but the policy depends on each model and provider.
Caching is most valuable for Agents or conversations that reuse the same System Prompt/Context across many turns.
See the Caching guide →

F68 AI Stocks is the first vertical AI Fintech of F68 AI.
The system combines:
F68 Data → F68 Intelligence → F68 Trading
with 5 specialist Agents for Technical, Fundamental, Macro, Combined and Investment Assistant.

A prompt is a request at one point in time. F68 Context systematizes knowledge, methods and principles so the Agent can use them throughout the analysis.

Data is extracted, standardized and organized into series before being fed to the AI, helping the Agent understand position, state and progression instead of just reading isolated numbers.

Not required. Choose an Agent and provide the necessary information, and the Agent will operate by the method and process already set up.

The Agent continues from the previous line of analysis, adds new data and assesses which theses or scenarios are being reinforced, weakened or need adjusting.

The Agent turns analysis results into an action plan based on position and portfolio status, including scenarios, confirmation conditions, invalidation conditions and risk management.

F68 supports research, analysis, and building scenarios and plans; it is not a personalized investment advisory service.

No. Both AI and data can contain errors or may not fully reflect new developments. For important decisions, users should check the data and weigh their goals and risk tolerance before acting.