Choosing the Right AI Model: Tiers, Access & Limits
This guide maps model capabilities, costs, and access rules in one place, so you can choose the exact tool along with the permission level that your task requires.
Important Disclaimer: AI models change rapidly and providers can add and remove models based on new updates and changes. Keep up to date by reading this topic.
Last updated - August 2025
1 · Model Tiers at a Glance
AI models in general vary in capabilities, cost structures, and use cases. Below is a structured breakdown of models and their tiers based on cost structure. Each tier has a specific usage limits associated with it to your user trust-level.
| Cost | Typical Use-Case | Example Models (Provider) |
|---|---|---|
 Low |
Quick drafts, data cleanup | llama-3.2-3b (Venice AI) |
 Mid |
Coding help, analysis | mistral-31-24b (Venice AI) |
 High |
Complex R&D, large context windows | qwen3-235b (Venice AI) |
All BIThub AI Providers
- Venice AI: Venice AI Model Pricing Full List
- OpenAI: ChatGPT Model Pricing Full List
- OpenRouter: OpenRouter Models Pricing Full List
- MorpheusAI: About MorpheusAI
All BIThub AI Models & Cost Tier
High Cost Mid Cost Low Cost Limited Trial deepseek-r1-671b deepseek-coder-v2-lite qwen3-4b DeepSeek R1 0528 llama-3.1-405b venice-uncensored (Dolphin Mistral) llama-3.2-3b DeepSeek V3 qwen3-235b qwen-2.5-coder-32b mistral-3.1-24b Llama 4 Maverick dolphin-2.9.2-qwen2-72b qwen-2.5-qwq-32b Llama 4 Scout llama-3.3-70b qwen-2.5-vl
All BIThub AI Models By Provider
Provider Models VeniceAI deepseek-coder-v2-lite,deepseek-r1-671b,dolphin-2.9.2-qwen2-72b,llama-3.2-3b,llama-3.3-70b,llama-3.1-405b,mistral-3.1-24b,qwen-2.5-coder-32b,qwen-2.5-qwq-32b,qwen-2.5-vl,qwen3-4b,qwen3-235b,venice-uncensoredOpenAI Test model: GPT-4.1 NanoOpenRouter Test Model: DeepSeek R1 0528,DeepSeek V3,Llama 4 Maverick,Llama 4 ScoutMorpheusAI llama-3.3-70b-web
- Test Models mean that we are testing to see if we do a full integration and/or we use them for admin tool backend testing.
2 · Four Ways to Interact with our AI tools
Table featuring main ways to use the
BITCORE AItoolsets.
| AI Mode | How does it work | Best For | Accessibility | Limits |
|---|---|---|---|---|
| 1. Discourse AI Chat | Private message an AI Bot using Discourse Private messages | Private inference with general tools. | Private to logged-in members only | All messages limited to 32 k chars. Each model has limits based on its price and costs. |
| 2. Discourse Category Workflows | Post a private topic that activates a workflow using the post as an input for the workflow | Advanced multi-response AI responses | Logged-in members can post in both Private & Public Categories. | Each CORE has its own set of rules and limits. |
| 3. AI Nodes | Each topic in the Nodes category hosts a CLI-like chat that powers a specific agent optimized for specific tasks | Power-user scripts / CLI | Non-Members can access certain free Nodes. Logged-in members can use more advanced Nodes. | Each Node has its own set of rules and limits. |
| 4. AI Workspaces | Experimental granular workspaces that use advanced workflows that are customizable to users tasks. | Advanced granular workflows | Non-Members can access certain free Workspaces. Logged-in members can use more advanced workspaces. | Each Workspace has its own set of rules and limits. |
Learn more on the Different Ways to Interact with our AI
Currently, there are 4 main ways to use our AI tools.
1. AI BOTs.
The first way is using the in-house Discourse AI tools that extend the Discourse Forum. The software that powers these tools get updated constantly and new features are added on a rolling basis.
Discourse Forum has an active software lifecycle and as such some feature might break and patches get added constantly. Additionally, the forum has all kinds of micro-features that help you with all kinds of things like translating, summarizing and helping you build posts.
What is this feature
You can send private messages to Constructs and other AI personas. Use advanced chat bots with specific constructs for inference and tools for all various kinds of general tasks.
Learn more about these Discourse features in their website here.
2. NODES.
Nodes are a special category that hosts CPU-like Terminal or Console-CLI interfaces that activate AI workflows. Each Node is standalone has its own use cases, limits and permissions. Some nodes can be very simple while other nodes can activate complex workflows and be restricted to certain user levels.
What is this feature
Find the Agentic Node that you want to activate in the Node Category and send queries. Nodes are modular and optimized and use specific tools for all kinds of modular tasks.
Learn more about NODES here.
3. CORES.
Core Constructs are advanced workflows that get activated when you post on the Cores category.
What is this feature
Post in CORE Categories and activate complex workflows and return multi-responses.
Learn more about these here.
4. Workspaces.
TBA
What is this feature
TBA
Learn more about these here.
3 · Access, Permissions and Limits
- We currently offer
publicendpoints and alsomember-onlyendpoints and AI tools.
- When members first log in they start at
Trust Level - 0. As users interact with the platform they start gaining XP to get to higher levels. Users can typically pass toTrust Level - 1on the same day. This is done to slowly roll AI tools to new users and onboard users to participate.
Access to AI models is restricted by user trust levels (T0–T3), determined by forum participation.
| Trust Level | Model Tiers Unlocked | Daily Calls |
|---|---|---|
| T-1 | All models | 1x calls |
| T-2 | All models | 2x calls |
| T-3 | All models | 3x calls |
Learn more about the trust levels and participation details in here.
AI Global Character Limit
Each prompt and each reply can be up to 32 000 characters.
PRIVACY NOTICE:
All prompts and completions for are logged and may be used to improve our products and services. You remain responsible for any required end user notices and consents and for ensuring that no personal, confidential, or otherwise sensitive information, including data from individuals under the age of 18, is submitted.
Learn more about the privacy features we offer.
Reference
Learn more on how Discourse Trust Levels work by visiting this page.