KLM 1.6
KLM 1.6 is an open-weights language model from MinimaLabs, released November 9, 2026. It has 0.1B parameters (dense). The context window is 2K tokens.
Check it against your own PC →
Lab
MinimaLabs
Released
November 9, 2026
Weights
Open apache-2.0
Parameters
0.1B dense
Context
2K
Price / M tokens
–
Epoch ECI
not yet rated
Can you run KLM 1.6 locally?
The smallest common GPU it runs on is 6 GB of VRAM (with 32 GB of RAM): Q6_K, fully on GPU, roughly 580 tokens per second. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 1.1 GB | – | ~580 |
| 8 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 1.1 GB | – | ~670 |
| 12 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 1.1 GB | – | ~720 |
| 16 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 1.1 GB | – | ~790 |
| 24 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 1.1 GB | – | ~870 |
| 32 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 1.1 GB | – | ~930 |
Benchmark results
Epoch AI has not published results for KLM 1.6 yet. New models are usually evaluated within days to weeks.