AI21 Jamba Reasoning 3B needs about 4.8 GB of memory at Q4_K_M (a 1.9 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 25 quants, from 0.0 GB to 6.4 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 9.4 GB.
AI21 Jamba Reasoning 3B – a 3B hybrid Transformer–Mamba model for long‑context reasoning
AI21 Jamba Reasoning 3B is a 3.2B parameter model that blends Transformer attention with Mamba state‑space layers. It supports up to 256k context tokens and is optimized for efficient sequence processing on laptops, GPUs and mobile devices. The model is released under the Apache‑2.0 license.
Good at
- Long‑context reasoning
- Efficient inference on limited VRAM
- Hybrid attention for complex dependencies
Limits
- No vision support
- Requires at least 4GB VRAM for Q4_K_M
- Context length limited to 256k tokens
Best for
- Edge deployment with long documents
- Datacenter workloads needing large context
- Reasoning tasks on low‑end GPUs
Who it’s for. Developers and researchers who need a compact model with strong reasoning and long‑context capabilities. The model is suitable for applications that require efficient inference on laptops or mobile devices while handling large inputs.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.0 GB | 2.9 GB | Fits | Get file → | |
| IQ2_M | 1.1 GB | 4.0 GB | Fits | Get file → | |
| Q2_K | 1.2 GB | 4.1 GB | Fits | Get file → | |
| IQ3_XXS | 1.3 GB | 4.1 GB | Fits | Get file → | |
| Q2_K_L | 1.4 GB | 4.2 GB | Fits | Get file → | |
| IQ3_XS | 1.4 GB | 4.3 GB | Fits | Get file → | |
| Q3_K_S | 1.5 GB | 4.3 GB | Fits | Get file → | |
| IQ3_M | 1.5 GB | 4.3 GB | Fits | Get file → | |
| Q3_K_M | 1.5 GB | 4.4 GB | Fits | Get file → | |
| Q3_K_L | 1.6 GB | 4.5 GB | Fits | Get file → | |
| Q3_K_XL | 1.8 GB | 4.6 GB | Fits | Get file → | |
| IQ4_XS | 1.8 GB | 4.6 GB | Fits | Get file → | |
| IQ4_NL | 1.9 GB | 4.7 GB | Fits | Get file → | |
| Q4_0 | 1.9 GB | 4.7 GB | Fits | Get file → | |
| Q4_K_S | 1.9 GB | 4.7 GB | Fits | Get file → | |
| Q4_K_M | 1.9 GB | 4.8 GB | Fits | Get file → | |
| Q4_1 | 2.0 GB | 4.9 GB | Fits | Get file → | |
| Q4_K_L | 2.1 GB | 4.9 GB | Fits | Get file → | |
| Q5_K_S | 2.2 GB | 5.1 GB | Fits | Get file → | |
| Q5_K_M | 2.3 GB | 5.1 GB | Fits | Get file → | |
| Q5_K_L | 2.4 GB | 5.2 GB | Fits | Get file → | |
| Q6_K | 2.6 GB | 5.5 GB | Fits | Get file → | |
| Q6_K_L | 2.7 GB | 5.6 GB | Fits | Get file → | |
| Q8_0 | 3.4 GB | 6.3 GB | Fits | Get file → | |
| BF16 | 6.4 GB | 9.3 GB | Too big | Get file → |
For 8GB VRAM, use Q4_K_M (file size 1.93GB, memory 4.8GB). If 6GB VRAM is available, Q3_K_M (1.54GB, 4.4GB) is a good balance. For 4GB VRAM, Q2_K (1.21GB, 4.1GB) is the lightest option. Memory estimates are approximate.
What GPU do you need for AI21 Jamba Reasoning 3B?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is IQ2_M (1.1 GB file, about 4.0 GB in use).
- 6 GB VRAM: the best quant that fits is Q6_K_L (2.7 GB file, about 5.6 GB in use).
- 8 GB VRAM: the best quant that fits is Q8_0 (3.4 GB file, about 6.3 GB in use).
- 12 GB VRAM: the best quant that fits is BF16 (6.4 GB file, about 9.3 GB in use).
Which AI21 Jamba Reasoning 3B quant should you choose?
Q4_K_M is the usual starting point for AI21 Jamba Reasoning 3B: it balances file size and quality well. Here it is a 1.9 GB file.
The Q8_0 file is 3.4 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is the first one in the table at 0.0 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about AI21 Jamba Reasoning 3B
How much VRAM does AI21 Jamba Reasoning 3B need?
About 4.8 GB at Q4_K_M with an 8K context window. That is the 1.9 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run AI21 Jamba Reasoning 3B on an 8 GB GPU?
Yes, with Q8_0: the 3.4 GB file needs about 6.3 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of AI21 Jamba Reasoning 3B GGUF?
It depends on the quant: file is 0.0 GB, IQ2_M is 1.1 GB, Q2_K is 1.2 GB, IQ3_XXS is 1.3 GB, Q2_K_L is 1.4 GB, IQ3_XS is 1.4 GB and 19 more.
What context length does AI21 Jamba Reasoning 3B support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 2.3 GB at 8K and 9.4 GB at 32K.
Where can I download AI21 Jamba Reasoning 3B GGUF?
The files are on Hugging Face in the bartowski/ai21labs_AI21-Jamba-Reasoning-3B-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does AI21 Jamba Reasoning 3B use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run AI21 Jamba Reasoning 3B on 8 GB VRAM?
Yes, using the Q4_K_M quantization which requires about 4.8GB of VRAM at 8k context.
Does the model support vision tasks?
No, the model only supports text and does not include vision capabilities.
What is the maximum context length I can use?
The model supports up to 256k tokens, but practical inference may be limited by available VRAM.