THOX Pro Model Catalog
Privacy-first AI models optimized for professionals across healthcare, legal, finance, research, and software engineering.
THOX.ai Pro models bring enterprise-grade AI to professionals who demand privacy, speed, and reliability. 100% local processing means your sensitive data never leaves your device — HIPAA, GDPR, and SOC2 compliant by design.
Healthcare
HIPAA Compliant
Legal
Attorney-Client Privilege
Enterprise
SOC2 & GDPR
Research
FERPA Compliant
Recommended for Most Users
Start with THOX Pro 8B — it offers vision, multilingual support across 32+ languages, and a 256K context window in just 10GB, making it ideal for most professional workflows on a single device.
THOX Pro Model Lineup
THOX Pro 8B
RecommendedProfessionalthox-pro-8b
Vision-enabled multilingual AI for professionals. Based on Ministral 3, supports 32+ languages with image analysis capabilities.
Size
8B parameters
Memory
10GB
Speed
40-60 tok/s
Context
256K tokens
Base Model
ministral-3:8b
Key Industries:
Key Capabilities:
- Vision & image analysis
- Multilingual (32+ languages)
- General professional AI
- Medical imaging
- Document visual review
THOX Pro 14B
Professionalthox-pro-14b
Advanced vision model with enhanced multilingual capabilities for complex professional tasks.
Size
14B parameters
Memory
14GB
Speed
30-45 tok/s
Context
256K tokens
Base Model
ministral-3:14b
Key Industries:
Key Capabilities:
- Advanced vision analysis
- Multilingual expertise
- Complex reasoning
- Visual document analysis
- Cross-language workflows
THOX Pro 24B
Enterprisethox-pro-24b
Agentic coding model for software engineering. Features tool use, function calling, and multi-file editing capabilities.
Size
24B parameters
Memory
20GB
Speed
25-35 tok/s
Context
384K tokens
Base Model
devstral-small-2:24b
Key Industries:
Key Capabilities:
- Agentic coding
- Tool use & function calling
- Codebase exploration
- Multi-file editing
- Software engineering workflows
THOX Pro 30B
Enterprisethox-pro-30b
Reasoning powerhouse with thinking mode. MoE architecture with 3.5B active parameters for advanced analysis.
Size
30B parameters (3.5B active)
Memory
24GB
Speed
35-50 tok/s
Context
256K tokens
Base Model
nemotron-3-nano:30b
Key Industries:
Key Capabilities:
- Advanced reasoning
- Thinking mode
- Complex analysis
- Mathematical reasoning (AIME 99.2%)
- PhD-level science (GPQA 73%)
THOX Pro 32B
Enterprisethox-pro-32b
Vision specialist with OCR for 32 languages. Excels at document-heavy workflows and spatial understanding.
Size
32B parameters
Memory
24GB
Speed
20-30 tok/s
Context
256K tokens
Base Model
qwen3-vl:32b
Key Industries:
Key Capabilities:
- Vision & OCR (32 languages)
- Document analysis
- Spatial understanding
- Visual reasoning
- Scanned document processing
Installing Models
# List available THOX.ai models on Ollama
$ ollama search THOX-ai
# Pull THOX Pro 8B (recommended)
$ ollama pull THOX-ai/thox-pro-8b
$ ollama run THOX-ai/thox-pro-8b
# For agentic coding (tool use, multi-file editing)
$ ollama pull THOX-ai/thox-pro-24b
$ ollama run THOX-ai/thox-pro-24b
# For advanced reasoning research
$ ollama pull THOX-ai/thox-pro-30b
$ ollama run THOX-ai/thox-pro-30b
# For document-heavy OCR workflows
$ ollama pull THOX-ai/thox-pro-32b
$ ollama run THOX-ai/thox-pro-32b
# Check installed models
$ ollama list
Quantization Options
Quality vs. Performance Tradeoffs
Each model is available in different quantization levels, trading off quality for speed and memory usage.
| Quantization | Quality | Speed | Memory (7B) |
|---|---|---|---|
| fp16 | Highest | Slowest | ~14GB |
| q8_0 | Very High | Fast | ~7GB |
| q4_k_m | Good | Very Fast | ~4GB |
| q4_0 | Acceptable | Fastest | ~3.5GB |
Recommendation: For most users, q4_k_m offers the best balance of quality and performance. Use q8_0 if you need higher quality and have memory to spare.
Model Selection Guide
| Use Case | Recommended | Memory | Why |
|---|---|---|---|
| Everyday professional AI | thox-pro-8b | 10GB | Vision + 32 languages, fastest tok/s |
| Complex visual analysis | thox-pro-14b | 14GB | Advanced vision + reasoning |
| Software engineering | thox-pro-24b | 20GB | Agentic coding, tool use, 384K context |
| Advanced research & math | thox-pro-30b | 24GB | Thinking mode, AIME 99.2%, GPQA 73% |
| Document OCR (32 languages) | thox-pro-32b | 24GB | Vision specialist, scanned documents |
Complete Privacy & Compliance
All THOX Pro models run entirely on your local THOX.ai device. No data is transmitted to external servers. Your medical records, legal documents, financial reports, and conversations never leave your device.
CONFIDENTIAL AND PROPRIETARY INFORMATION
This documentation is provided for informational and operational purposes only. The specifications and technical details herein are subject to change without notice. THOX.ai LLC reserves all rights in the technologies, methods, and implementations described.
Nothing in this documentation shall be construed as granting any license or right to use any patent, trademark, trade secret, or other intellectual property right of THOX.ai LLC, except as expressly provided in a written agreement.
Patent Protection
The MagStack™ magnetic stacking interface technology is proprietary technology of THOX.ai LLC, protected by trade secrets and intellectual property laws....
Reverse Engineering Prohibited
You may not reverse engineer, disassemble, decompile, decode, or otherwise attempt to derive the source code, algorithms, data structures, or underlying ideas of any THOX.ai hardwa...
THOX.ai™, ThoxOS™, MagStack™, MeshStack™, ThoxMigrate™, the THOX Edge Series™, the THOX Nova Series™, and the THOX.ai logo are trademarks or registered trademarks of THOX.ai LLC in the United States and other countries. WireGuard® is a registered trademark of Jason A. Donenfeld.
All other trademarks are the property of their respective owners.
© 2026 THOX.ai LLC. All Rights Reserved.