tiny aya l2 thinker
tiny aya l2 thinker is an open-weights language model from Cohere, released September 2, 2026. It has 3.4B parameters (dense).
Check it against your own PC →
Lab
Cohere
Released
September 2, 2026
Weights
Open
Parameters
3.4B dense
Context
–
Price / M tokens
–
Epoch ECI
not yet rated
Can you run tiny aya l2 thinker locally?
The smallest common GPU it runs on is 6 GB of VRAM (with 32 GB of RAM): Q6_K, fully on GPU, roughly 40 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) | 5.1 GB | – | ~40 |
| 8 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 5.1 GB | – | ~56 |
| 12 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 5.1 GB | – | ~73 |
| 16 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 5.1 GB | – | ~98 |
| 24 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 5.1 GB | – | ~170 |
| 32 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 5.1 GB | – | ~280 |
Benchmark results
Epoch AI has not published results for tiny aya l2 thinker yet. New models are usually evaluated within days to weeks.
Links
Hugging Face: CohereLabs/tiny-aya-l2-thinker
More from Cohere
tiny aya en thinker · North Micro Vision Instruct · North Mini Code · Command A Plus · Tiny Aya Global · Tiny Aya Fire · Tiny Aya Water · Tiny Aya Earth