Models we advise on
Picked from the world agent leaderboard, filtered to open-source (MIT / Apache / open licenses), all around or above 100 billion parameters — the scale where a serious cluster is required.
MIT
Full breakdown
GLM 5.2 Max
Z.AI
A frontier-scale open-weight reasoning and agent model from Z.AI, released under the permissive MIT license.
| Parameters | 744 billion |
|---|---|
| Minimum GPU memory | 892.8 GB |
| Cheapest working build | 12x NVIDIA H100 SXM ($384,000) |
MIT
Full breakdown
DeepSeek V4 Pro
DeepSeek
DeepSeek's largest open-weight model, built for top-tier agentic and coding performance, released under the permissive MIT license.
| Parameters | 1,600 billion |
|---|---|
| Minimum GPU memory | 1920.0 GB |
| Cheapest working build | 24x NVIDIA H100 SXM ($768,000) |
OpenMDW
Full breakdown
Nemotron 3 Ultra
NVIDIA
NVIDIA's flagship open-weight model, tuned for enterprise agent workloads, released under NVIDIA's Open Model Development Warranty license.
| Parameters | 550 billion |
|---|---|
| Minimum GPU memory | 660.0 GB |
| Cheapest working build | 9x NVIDIA H100 SXM ($288,000) |