Compare LLM Models
Select 2-5 models to compare side by side.
Select models to compare
0 / 5 selected
Select 2-5 models
| Attribute |
tokyotech-llm
|
|---|---|
| Company | tokyotech-llm |
| Release Date | 2026-02-20 |
| Parameters (B) | 32 |
| Architecture | Transformer |
| Context Window | 32768 |
| Input Modalities | text |
| Open Source | Yes |
| License | |
| Score | 8.4 |
| VRAM (GB) | 24 |
| Compute | Single NVIDIA GPU with 24GB+ VRAM (e.g. RTX 4090, RTX 3090, A5000); AWQ-INT4 weights ~18GB plus KV cache overhead; deploy via vLLM or SGLang with --reasoning-parser qwen3 |
| Benchmark: GSM8K | 98.2 |
| Benchmark: HumanEval | 63.8 |
| Benchmark: MMLU | 77.8 |
| Pricing (Input) | — |
| Pricing (Output) | — |