Marketplace
Everything in the launch catalog, with presale pricing. Listings marked “Coming soon” are in review and will open shortly after launch.
Model APIs
OpenAI-compatible chat completions. Billed per 1K tokens (input + output).
Multilingual instruction model covering 29+ languages including Urdu. Bilingual support chat, document summaries and form-filling help.
- Params
- 1.54B
- Context
- 32K
- Licence
- Apache-2.0
Small, fast model for intent routing, ticket triage, short JSON extraction and guard-rail checks in front of larger models.
- Params
- 0.49B
- Context
- 32K
- Licence
- Apache-2.0
Hybrid conv/attention architecture tuned for fast inference. Structured extraction, RAG answer synthesis and tool-calling agents.
- Params
- 350M
- Context
- 32K
- Licence
- LFM Open License v1.0
Compact English-first model for low-bandwidth deployments: FAQ answering, rewriting and short summaries.
- Params
- 360M
- Context
- 8K
- Licence
- Apache-2.0
Llama-2 architecture trained on ~3T tokens and aligned with UltraChat + UltraFeedback. A solid baseline for helpdesk chat.
- Params
- 1.1B
- Context
- 2K
- Licence
- Apache-2.0
Fine-tuned for Urdu news categories: business, entertainment, sports, science-technology, world and Pakistan.
- Params
- 1.54B
- Context
- 32K
- Licence
- CC-BY-4.0 data
GPU capacity
Billed per GPU-hour. Provider-run capacity is listed by third parties and labelled on every job.
Want a model or GPU that isn't listed?
Tell us in the presale form and we'll prioritise it for launch.