IB

IBM Grok4 UltraFast Coder 1B

IbmGGUF by mradermacherUpdated 2026-04-22
Download on Hugging Face
Parameters1.63B
Context128K
Layers40
Architecturegranite
Smallest quant0.7 GB
Quick answer

IBM Grok4 UltraFast Coder 1B needs about 4.2 GB of memory at Q4_K_M (a 1.0 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 12 quants, from 0.7 GB (Q2_K) to 3.3 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 10.7 GB.

IBM Grok4 UltraFast Coder 1B – a compact coding assistant

IBM Grok4 UltraFast Coder 1B is a 1.63 billion‑parameter granite model focused on coding and tool usage. It supports a large context of 131072 tokens and is available in multiple GGUF quantizations.

Good at

  • Coding assistance
  • Tool integration
  • Large context handling

Limits

  • No vision support
  • License not specified
  • No mixture of experts

Best for

  • Writing code snippets
  • Debugging help
  • Automating coding tasks

Who it’s for. Developers and researchers needing a lightweight coding model that can run on modest hardware. The model is suitable for local deployment with GGUF support.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K0.7 GB 3.9 GB
Fits Get file →
Q3_K_S0.8 GB 4.0 GB
Fits Get file →
Q3_K_M0.9 GB 4.0 GB
Fits Get file →
Q3_K_L0.9 GB 4.1 GB
Fits Get file →
IQ4_XS0.9 GB 4.1 GB
Fits Get file →
Q4_K_S1.0 GB 4.2 GB
Fits Get file →
Q4_K_M1.0 GB 4.2 GB
Fits Get file →
Q5_K_S1.2 GB 4.3 GB
Fits Get file →
Q5_K_M1.2 GB 4.4 GB
Fits Get file →
Q6_K1.3 GB 4.5 GB
Fits Get file →
Q8_01.7 GB 4.9 GB
Fits Get file →
F163.3 GB 6.5 GB
Fits Get file →

For 8 GB VRAM, use Q4_K_M (1.02 GB file, ~4.2 GB memory). If 4 GB VRAM is available, Q3_K_S (0.79 GB file, ~4 GB memory) is the smallest that fits. For higher quality, Q5_K_M (1.18 GB file, ~4.4 GB memory) can be used on 8 GB VRAM.

What GPU do you need for IBM Grok4 UltraFast Coder 1B?

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 Q3_K_M (0.9 GB file, about 4.0 GB in use).
  • 6 GB VRAM: the best quant that fits is Q8_0 (1.7 GB file, about 4.9 GB in use).
  • 8 GB VRAM: the best quant that fits is F16 (3.3 GB file, about 6.5 GB in use).

Which IBM Grok4 UltraFast Coder 1B quant should you choose?

Q4_K_M is the usual starting point for IBM Grok4 UltraFast Coder 1B: it balances file size and quality well. Here it is a 1.0 GB file.

The Q8_0 file is 1.7 GB, about 1.7 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 Q2_K at 0.7 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturegranite
Parameters1.63B
Layers40
Trained context131,072 tokens
KV cache (fp16)320 KB per token

Source

Quantized bymradermacher
FamilyIbm
Downloads (HF)815
Last updated2026-04-22

Frequently asked questions about IBM Grok4 UltraFast Coder 1B

How much VRAM does IBM Grok4 UltraFast Coder 1B need?

About 4.2 GB at Q4_K_M with an 8K context window. That is the 1.0 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 IBM Grok4 UltraFast Coder 1B on an 8 GB GPU?

Yes, with F16: the 3.3 GB file needs about 6.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of IBM Grok4 UltraFast Coder 1B GGUF?

It depends on the quant: Q2_K is 0.7 GB, Q3_K_S is 0.8 GB, Q3_K_M is 0.9 GB, Q3_K_L is 0.9 GB, IQ4_XS is 0.9 GB, Q4_K_S is 1.0 GB and 6 more.

What context length does IBM Grok4 UltraFast Coder 1B support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 2.7 GB at 8K and 10.7 GB at 32K.

Where can I download IBM Grok4 UltraFast Coder 1B GGUF?

The files are on Hugging Face in the mradermacher/IBM-Grok4-UltraFast-Coder-1B-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Can I run this model on 8 GB VRAM?

Yes, the Q4_K_M quant uses about 4.2 GB of VRAM at 8 k context, fitting within an 8 GB limit.

Does this model support image input?

No, the model has no vision capabilities.

What is the context length of this model?

It supports up to 131072 tokens of context.

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Ibm models

All Ibm →

Other tiny models

All tiny models →