👋 Join our WeChat or Discord community. 📖 Check out the GLM-5.3-Flash blog and GLM-5 Technical report . 📍 Use GLM-5.3-Flash API services on Z.ai API Platform.
Inkling-2-Small
thinkingmachines·Instruct·released 4 Nov 2026
ollama run inkling-2-smallInkling 2 Small is a general-purpose multimodal model that accepts text and image inputs and generates text outputs. It is intended for use in English and other languages, and across multiple coding languages.
How much VRAM does Inkling-2-Small need?
At Q4_K_M Inkling-2-Small needs about 229 GB with an 8K context (229 GB at Q4, 751 GB at full precision). Pick your GPU to see what fits and how fast it runs.
Needs: ≈229 GB · Q4_K_M · 8K context — Pick a GPU to check the fit.
Advanced: context 8K · KV cache 16-bit
Smallest GPU for each quant and context (KV cache 16-bit)
| Quantisation | 8K | 16K | 32K | 64K | 128K | 1M |
|---|---|---|---|---|---|---|
| Q3_K_M | ||||||
| Q4_K_M | ||||||
| Q5_K_M | ||||||
| Q6_K | ||||||
| Q8_0 | ||||||
| FP16 |
≤16 GB card24–32 GB48–96 GBmore
| Quantisation | Total |
|---|---|
| Q2_KHeavy lossSmallest, noticeably worse quality | ≈125 GB |
| Q3_K_MNoticeable lossSmall, some quality loss | ≈185 GB |
| Q4_K_MrecommendedSmall lossRecommended default for local use | ≈229 GB |
| Q5_K_MNear-losslessNear-lossless for most tasks | ≈268 GB |
| Q6_KNear-losslessPractically lossless | ≈311 GB |
| Q8_0LosslessLossless in practice, twice the size of Q4 | ≈400 GB |
| FP16 / BF16LosslessFull precision, as released | ≈476 GB |
Estimates for 8K tokens of context: weights + KV cache + runtime overhead. All experts stay in memory, but only the active ones are read per token. Speed assumes memory-bound decoding at your GPU’s bandwidth. Pick a GPU above and the fit column checks every row against it. Weights marked “file” use the real GGUF download size; “est.” rows are computed from the parameter count.
How to pick a quant
Take the biggest quant that fits with room left for context. Q4_K_M is the usual sweet spot. Below Q4 quality drops fast, above Q6 you mostly pay memory for nothing. Big models (70B+, large MoE) tolerate low quants better than small ones.
- LosslessF16, Q8_0
- Near-losslessQ6_K, Q5_K_M
- Small lossQ5_K_S, Q4_K_M, IQ4_NL, IQ4_XS, Q4_K_S
- Noticeable lossQ4_0, Q3_K_L, Q3_K_M, IQ3_M
- Heavy lossIQ3_XXS, Q3_K_S, IQ2_M, Q2_K
Typical quality loss, not measured per model. Tiers come from KL divergence against FP16 on Llama 3 8B: llama.cpp perplexity README: LLaMA 3 8B scoreboard (KLD and ΔPPL vs FP16, Wikitext-2); llama.cpp quantize tool: ΔPPL per quant type @ Llama-3-8B
How much context fits?
At 8K context, Q4_K_M needs ≈229 GB(216 GB weights + 1.5 GB KV cache + 11 GB overhead). Pick a GPU above to see how much context fits.
Weights + overheadKV cache
Can the Apple M3 Ultra (512 GB) run Inkling-2-Small? →
Speed vs quality per quant
Smaller quants run faster and fit in less memory, but lose quality.
Fits in VRAMTight fitCPU offload✕ won't runQuality is typical KL divergence vs FP16, not measured for this model.
Real-world speed of Inkling-2-Small
Tokens per second measured by people running it at home. A GPU shows a median once 3 reports agree on the same setup. Reports far from our estimate wait for a review. Measure yours with llama-bench →
No medians yet. Be one of the first 3 to report a setup.
All reports
Run Inkling-2-Small locally
--quantQ4_K_Mdefault; pick your GPU to tailor itunsloth/Inkling-2-Small-GGUF
ollama run inkling-2-smallThe library build picks its own default quantisation.
ollama run hf.co/unsloth/Inkling-2-Small-GGUF:UD-Q4_K_XLfrom transformers import pipeline
pipe = pipeline("image-text-to-text", model="thinkingmachines/Inkling-2-Small", dtype="auto", device_map="auto")
messages = [
{
"role": "user",
"content": [
{"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/bee.jpg"},
{"type": "text", "text": "Describe this image."},
],
}
]
out = pipe(text=messages, max_new_tokens=256)
print(out[0]["generated_text"][-1]["content"])Quantised versions and downloads
| Repository | Format | Quant | Size |
|---|---|---|---|
| btbtyler09/Inkling-2-Small-NVFP4 | Safetensors | NVFP4 | — |
| Jon-Nielsen/Inkling-2-Small-NVFP4-CSF-FLARE-sm120-1M-vision | Safetensors | NVFP4 | — |
| unsloth/Inkling-2-Small-GGUF | GGUF | UD-IQ1_S | 172 GB |
| unsloth/Inkling-2-Small-GGUF | GGUF | UD-IQ1_M | 179 GB |
| unsloth/Inkling-2-Small-GGUF | GGUF | UD-IQ2_XXS | 191 GB |
| unsloth/Inkling-2-Small-GGUF | GGUF | UD-Q2_K_XL | 205 GB |
| unsloth/Inkling-2-Small-GGUF | GGUF | UD-IQ3_XXS | 226 GB |
| unsloth/Inkling-2-Small-GGUF | GGUF | UD-Q3_K_XL | 285 GB |
| unsloth/Inkling-2-Small-GGUF | GGUF | UD-Q4_K_XL | 354 GB |
| unsloth/Inkling-2-Small-GGUF | GGUF | UD-Q8_K_XL | 387 GB |
About Inkling-2-Small
Inkling 2 Small is a general-purpose multimodal model that accepts text and image inputs and generates text outputs. It is intended for use in English and other languages, and across multiple coding languages. The model is designed to be used by developers building AI-powered applications, including agentic and tool-use systems, coding assistants, chatbots, and retrieval-augmented generation systems, and is suitable for general-purpose conversational use, instruction-following, and other natural language and visual tasks. It is released with open weights to support research, fine-tuning and integration into third-party products by downstream developers. Languages: English, with general multilingual capabilities across other languages. Try Inkling 2 Small on the Tinker Playground or access via API using the Tinker Cookbook. To fine-tune Inkling 2 Small with Tinker, also install the tml-renderers package. Inkling 2 Small supports local inference using: • SGLang (recipe) • vLLM (recipe) • TokenSpeed (recipe) • Unsloth (recipe) Inkling 2 Small can also be post-trained with: • Miles • SkyRL • Snowflake Cortex Training API access is also available through third party inference providers, including: • NVIDIA • Fireworks • Together • Baseten • Nebius • Modal • Cloudflare The model is distributed in two checkpoint formats.
Full model card on Hugging Face →Alternatives to Inkling-2-Small
All alternatives →Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
MiniMax-M3 is a native multimodal model with 1M context. It has ~428B parameters and ~23B activated parameters.
Inkling-Small is a general-purpose multimodal model that accepts text, image and audio inputs and generates text outputs. It is intended for use in English and other languages, and across multiple coding languages.
👋 Join our WeChat or Discord community. 📖 Check out the GLM-5.3-Flash blog and GLM-5 Technical report . 📍 Use GLM-5.3-Flash API services on Z.ai API Platform.
We are excited to introduce DeepSeek-V4-Flash-Vision-Exp, our first experimental multimodal model in the DeepSeek-V4 family. It builds on the DeepSeek-V4-Flash architecture by incorporating visual modules and undergoing continued training to unlock visual understanding capabilities.
Inkling-2-Small FAQ
- How much VRAM does Inkling-2-Small need?
- About 229 GB at Q4_K_M with an 8K context. Q8_0 needs about 400 GB and full precision about 751 GB. Longer contexts need more memory for the KV cache.
- Can Inkling-2-Small run on the Apple M3 Ultra (512 GB)?
- Yes. It fits at Q6_K (about 311 GB), at an estimated 1.4 tokens/s.
- Can I use Inkling-2-Small commercially?
- Yes. It is released under Apache 2.0. Permissive: use, modify and sell freely; keep the notice and licence text.
- What is the context length of Inkling-2-Small?
- 1,048,576 tokens (1M). The full window needs far more memory than an 8K context.
- How do I run Inkling-2-Small with Ollama?
- Run ollama run inkling-2-small in a terminal.
Rate this model
Data from Hugging Face, the model card and public leaderboards. Memory and speed figures are estimates; real numbers depend on the runtime, context length and drivers. Always check the licence yourself before commercial use.
0 comments