Jais 3 32B
Jais 3 32B is the largest dense member of Jais 3. Its model card advertises a 512K context window and links FP8, GGUF and MLX formats. It is published under Apache 2.0.
Local formats and the Uno adapter
The optional 32B Uno release adds a diffusion adapter to the frozen base. Its card gives a separate inference-engine installation and explicitly distinguishes the adapter from the complete model.Jais 3 32B Uno official adapter card
Compare dense execution with sparse alternatives
Practical guidance. Compare this release with MoVA on the same documents, tool calls and concurrency. Dense and sparse parameter counts describe different execution patterns. They do not establish a direct speed or quality ranking.Jais 3 32B official model card
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | Jais-3-32B |
| Architecture | Dense decoder-only; 32B classJais 3 32B official model card |
| Published context | 512K; developer notationJais 3 32B official model card |
| Weight license | Apache 2.0Jais 3 32B official model card |
Related models
- Jais 3 7B
Inception42 (G42) · Jais 3
Seven-billion-parameter Arabic-and-English model with reasoning and tool-call support.
Launch date not confirmedDownloadable - Jais 3 MoVA 36B-A4B
Inception42 (G42) · Jais 3
Arabic-and-English mixture-of-experts model combining 36 billion stored parameters with Mixture-of-Values attention.
Launch date not confirmedDownloadable