Can I run Inkling 2 Small locally?
Inkling 2 Small by thinkingmachines has 594B parameters (19B active per token, 384 experts, 8 used per token). At 4-bit with an 8K context it needs about 346 GB of GPU memory. It fits on 1 of 65 GPUs and Macs we track.
Memory needed by quant and context length
GB of GPU memory = weights + KV cache (FP16) + runtime overhead. Weights use real file sizes from bartowski/Inkling-2-Small-GGUF where available (marked •), otherwise parameter count × bits per weight. How we calculate.
| Quant | Quality | 4K ctx | 8K ctx | 32K ctx | 128K ctx |
|---|---|---|---|---|---|
| 2-bit (Q2_K) • | rough | 212 GB | 213 GB | 217 GB | 235 GB |
| 3-bit (Q3_K_M) • | usable | 262 GB | 263 GB | 268 GB | 286 GB |
| 4-bit (Q4_K_M) • | good, most popular | 345 GB | 346 GB | 350 GB | 368 GB |
| 5-bit (Q5_K_M) • | very good | 425 GB | 426 GB | 430 GB | 448 GB |
| 6-bit (Q6_K) • | near lossless | 488 GB | 489 GB | 493 GB | 511 GB |
| 8-bit (Q8_0) • | effectively lossless | 589 GB | 590 GB | 595 GB | 613 GB |
| 16-bit (F16/BF16) | original weights | 1108 GB | 1109 GB | 1113 GB | 1131 GB |
Which GPUs and Macs run Inkling 2 Small?
Best quality quant that fits at 8K context. "Tight" = within 10% of usable memory. Speed is a ceiling from memory bandwidth (real speed is typically 50–80% of it).
Prefer an API?
Thinking Machines: Inkling 2 Small is on OpenRouter at $0.30 per 1M input tokens and $1.00 per 1M output tokens. Price details and history →
Questions
How much VRAM does Inkling 2 Small need?
About 346 GB at 4-bit (Q4_K_M) with an 8K context: 344 GB of weights (real GGUF file size), 1.5 GB of KV cache and 0.8 GB of runtime overhead. At 8-bit it needs about 590 GB.
What is the smallest GPU that runs Inkling 2 Small?
No single consumer GPU in our table fits it at Q4_K_M with 8K context.
Can I run Inkling 2 Small on a Mac?
Not at Q4_K_M with 8K context on any Mac in our table, even a 512 GB Mac Studio, using the default GPU memory limit.
Model facts
| Hugging Face repo | thinkingmachines/Inkling-2-Small |
| Parameters | 594,050,672,720 (594B) · ~19B active (computed from config) |
| Layers / hidden size | 48 / 4608 |
| Attention | GQA; 48 heads, 8 KV heads, head dim 128 |
| KV cache per token (FP16) | 192.0 KiB |
| Max context | — |
| License | apache-2.0 |
| Released on HF | Nov 4, 2026 |
| Downloads (30d) / likes | 16,918 / 420 |
Sources: Hugging Face model API · config.json (thinkingmachines/Inkling-2-Small) · GGUF files · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.