TwIL 8B balanced
8.8BwebAI TwIL
A reasoning model for maths, code and instruction following, post-trained from IBM Granite 4.2 8B. This entry is the balanced checkpoint only; verbose is a distinct set of weights and its files are not mixed into these levels. All 40 layers keep a growing KV cache. The webAI licence permits non-commercial research and education only. Quantization quality loss has not been published.
131K
Max Context
1
Quant Variants
GGUF Q4_K_M
Best Quality
—
Perplexity rise (best build)
Quantization Variants
Per-quant VRAM, quality loss, and inference speed on RTX 4090
Measured = run on this site · Estimated = not run here: calculated, or a figure this site has not verified · Community = public reports
Which file each size is for
A file size is not a memory figure: the VRAM column is the calculator's total at 4K context — weights plus KV cache plus a 10% activation buffer. Units differ on purpose: file sizes are decimal GB, as Hugging Face lists them; the VRAM column is GiB (1024³ bytes) with 2% added to the weights, so 7.12 GB of file counts as about 6.8 GiB there. A vision projector (mmproj), where the repo ships one, is a separate download and is in neither number.
- GGUF Q4_K_MwebAI-Official/TwIL-8B · TwIL-8B-Q4_K_M-balanced.gguf · 5.35 GB (5,347,917,568 bytes) · listing read 2026-11-15
Related community runs
These runs describe their own settings. Their speeds do not replace this estimate or prove your configuration fits.
Our reviewed link index has no run for this model yet. You can browse community reports or share your own test below.
Browse more runs on llamaPerf ↗Similar models
Compare with Qwen3.5 9BQwen3.5 9B
Alibaba Qwen3.5
The newest small Qwen and a common pick for 8 to 16 GB cards. Only 8 of its 32 layers keep a growing KV cache, so it can use a long context on modest hardware. At Q4_K_M it needs about 6.6 GB at 4K and 7.0 GB at 16K, comfortable on an 8 GB card; at 32K (7.5 GB) it is tight there and comfortable from 10 GB up. Its full 256K window needs about 15.2 GB. Q8_0 needs about 10.4 GB, comfortable from 12 GB up. Reasoning is off by default on the small Qwen3.5 sizes, and it reads images through a separate vision encoder. Sizes come from the published GGUF file sizes, and no per-level quality loss has been published.
Granite 4.2 8B
IBM Granite
IBM's dense 8B reasoning model from the Granite 4.2 family (3B, 8B, 30B), released in August under Apache 2.0. Thinking is on by default; a non-thinking and a low-effort mode are chosen through the chat template. Unlike Granite 4.0's hybrid models it is a plain transformer, so llama.cpp runs it as `granite` and every one of its 40 layers keeps a full cache: about 6.5 GB at Q4_K_M with a 4K context, 11.3 GB at 32K and 27.8 GB across the whole 128K window. That makes it an 8 GB card model at short contexts and a 16 GB one for long documents. Sizes come from bartowski's GGUF files; IBM publishes no perplexity by quantization level, so no quality-loss figure is shown.
Qwen2.5 14B Instruct
Alibaba Qwen2.5
The sweet spot between performance and resource usage. 16GB VRAM with Q4.
Phi-3 Medium 14B Instruct
Microsoft Phi
Microsoft's mid-size Phi-3. Excellent quality-per-GB on 16GB cards.
Running TwIL 8B balanced locally
At Q4_K_M and 4K of context, TwIL 8B balanced needs about 6.3 GB — 5.1 GB of weights, 0.63 GB of KV cache and a 0.6 GB activation buffer. The smallest cards in this index that clear that comfortably are 8 GB ones — 9 of them, from the RTX 4060 up — and 76 of the 78 cards here do. These are calculated figures, not measurements: the estimate stops counting a card as comfortable at 88% of its VRAM, which is roughly the room a desktop session needs.
What longer context costs
Going from 4K to 32K adds about 4.37 GB, taking the total to 11.1 GB. Weights do not move with context — only the KV cache does, and it grows linearly, so this is the number to watch when planning for long documents. The model's native window is 128K; holding all of it at Q4_K_M would need about 28 GB.
Which build to download
This index only tracks TwIL 8B balanced in GGUF, across 1 level (Q4_K_M). No per-level perplexity sweep has been published for this model, so the quality column is empty rather than estimated — within one model, more bits per weight is the only ordering the data supports. No throughput has been measured for this model on this site's hardware. GGUF runs on llama.cpp and Ollama across NVIDIA, AMD and Apple silicon; AWQ and GPTQ target vLLM on CUDA and ROCm; EXL2 is ExLlamaV2 — CUDA only, and now archived, so local use rather than serving.
Common questions
- How much VRAM does TwIL 8B balanced need?
- About 6.3 GB at Q4_K_M with 4K of context and batch 1, rising to roughly 11.1 GB at 32K. That figure is weights plus KV cache plus a 10% activation buffer, calculated from the model's architecture rather than measured on a card.
- Will TwIL 8B balanced run on a 8GB GPU?
- Yes — at Q4_K_M and 4K context it needs about 6.3 GB, which leaves 1.7 GB spare on a RTX 4060, or on any of the other 8 GB cards here (9 in all). That is the smallest size in this index that clears it comfortably; 76 of 78 do.
- Which quantization of TwIL 8B balanced should I use?
- Q4_K_M is the only level in the index for this model, so there is nothing to choose between here.
Where this model fits
Sized at Q4_K_M with a 4K context window, smallest card first. Comfortable means the estimate uses at most 88% of the memory.
- Runs comfortably on
- RTX 40608GB · +1.7RTX 4060 Ti 8G8GB · +1.7Radeon RX 9060 XT 8G8GB · +1.7RTX 5060 Ti 8G8GB · +1.7RTX 50608GB · +1.7
- Tight but possible
- Mac M1 8G8GBMac M3 8G8GB
- Ships in
- GGUF
- Try it yourself
- Size it in the calculatorGenerate the run command
How to actually run this
Deployment guides for this model and this class of hardware.
This page's figures change when the model or the runtime does.Last updated 2026-11-15 RSS → /feed.xml