Llama-3.1-Tulu-3-8B
Ai2- Arena score
- 1193
- Among open models
- #112of 148
- Among all on Arena
- #327
- Votes
- 2,896
Measured on Llama 3.1 Tulu 3 8B, the served version, not a local build. LMArena, CC BY 4.0, published Nov 11, 2026.
Score by category
CategoryScoreOpen rankArena rankVotes
- English1215#113#3271.6K
- Business & finance1196#107#312285
- Math1195#106#312363
- Science1193#112#326516
- Russian1193#110#319418
- Applied math1193#107#314304
- Software & IT1191#114#329761
- Coding1183#114#330476
- Creative writing1181#109#322472
- Long prompts1181#113#331454
- Hard prompts, English1178#113#333452
- Writing & literature1178#112#330816
- Chinese1176#111#324238
- Hard prompts1174#113#328728
- Instructions1174#113#3271.2K
- Non-English1169#113#3281.3K
- Media & sports1161#110#325563
- Multi-turn1154#116#336520
- Excluding ties1060#113#3291.8K
| Model | Params | Score |
|---|---|---|
| 8.0B | 1193 | |
| 70.6B | 1256 |
About as strong, smaller
Compare
granite-4.2-3bIBM Granite12923.7B
Built on it
Ways to run it
- GGUF · llama.cpp, Ollama, LM Studiolmstudio-community/Llama-3.1-Tulu-3-8B-GGUF16 builds
- MLX · Apple silicon: mlx-lm, LM Studiomlx-community/Llama-3.1-Tulu-3-8B-8bit7 builds
- AWQ · vLLM, SGLangtayaee/Llama-3.1-Tulu-3-8B-AWQ-4bit1 build
Most downloaded
- Llama-3.1-Tulu-3-8B-GGUFlmstudio-community · Quantization5.4K downloads
- Llama-3.1-Tulu-3-8B-GGUFbartowski · Quantization1.4K downloads
- Llama-3.1-Tulu-3-8B-i1-GGUFmradermacher · Quantization765 downloads
- Llama-3.1-Tulu-3-8B-GGUFmradermacher · Quantization594 downloads
- allenai.Llama-3.1-Tulu-3-8B-GGUFDevQuasar-7 · Quantization456 downloads
BenchClub runs
Not runnable yet
Runs open once a Llama-3.1-Tulu-3-8B build is registered.
More from Ai2
See more
Molmo2-8B47.6K8.7B

Olmo-3.1-32B-Instruct21.3K32.2B

Olmo-3-32B-Think11.8K32.2B

Olmo-3.1-32B-Think7.7K32.2B
Other 3–9B models
See more
NVIDIA-Nemotron-3-Nano-4B-BF16NVIDIA1.9M4.0B
granite-4.1-8bIBM Granite79.4K8.8B
Spark-X2.6-4BXHToken32.9K4.1B
tiny-aya-globalCohere Labs31.7K3.3B