Leaderboard / PleIAs
Baguettotron
PleIAs · Pleias · released 2025-11-10 · data checked 2026-11-01
Baguettotron is a 320M-parameter dense model from PleIAs (Europe), released 2025-11-10 under the Apache-2.0 license. It doesn't have enough reported benchmarks for an index score yet. It needs about 0.9 GB of memory at 4-bit with an 8K-token context.
321M-parameter, 80-layer reasoning model trained only on the open SYNTH dataset.
Benchmarks
Each bar shows this model's score. The tick marks the best tiny model on that benchmark.
No benchmark results for this release yet. Know of one? Send it.
Specifications
- Lab
- PleIAs (FR · Europe)
- Architecture
- Dense transformer
- Total parameters
- 320M
- Active parameters
- All (dense)
- Size class
- Tiny
- Context window
- 4K tokens
- Input modalities
- text
- Reasoning
- Reasoning (thinking)
- License
- Apache-2.0
- Openness
- Fully open: weights, training data and code
- Memory to run
- ~0.9 GB at 4-bit, 8K context (more sizes)
- Hugging Face
- 2K downloads (30 days) · 268 likes
Compare with
- Baguettotron vs LFM2.5 350M (350M, index 12.6)
- Baguettotron vs Qwen3.5 0.8B (900M, index 10.7)
- Baguettotron vs Gemma 3 1B (1B, index 6.1)
Sources
Something wrong here? Report it.
Run it locally
Uses PleIAs/Baguettotron-GGUF (Q4_K_M). Swap the tag for another size: BF16, F16, IQ4_NL, IQ4_XS, Q4_0, Q4_1, Q4_K_M, Q4_K_S….
ollama run hf.co/PleIAs/Baguettotron-GGUF:Q4_K_Mllama-server -hf PleIAs/Baguettotron-GGUF:Q4_K_Mvllm serve PleIAs/BaguettotronMemory needed: weights, context cache and runtime buffers.
| Quantization | 4K context |
|---|---|
| 16-bit F16 | 1.4 GB |
| 8-bit Q8_0 | 1.1 GB |
| 4-bit Q4_K_M | 0.9 GB |