Model requests
Terminal can support additional open-source serverless models by request. Examples include:| Model family | Typical use |
|---|---|
| GLM 5.2 | General reasoning, coding, and long-context tasks |
| DeepSeek V4 Pro | General reasoning, coding, and agentic workloads |
| Kimi K2.6 | Coding, agentic workflows, and long-context tasks |
| DeepSeek V4 Flash | Fast, low-cost long-context inference |
| NVIDIA Nemotron 3 Ultra | Agentic and high-throughput production workloads |
Request a model
Email shubham@terminal.fyi with:- The model name or public model link.
- Your expected input and output token volume.
- Whether you need Standard, Flex, or both.
- Any latency or throughput target you care about.