whatmodel.sh / Jamba 1.6 Mini
Jamba 1.6 Mini
↗ share
+ contribute
Jamba 1.6 Mini is an open-weights language model from AI21 Labs, released February 2025. Its published size is 51.57B parameters (12B active per token). Licence: other (jamba-open-model-license).
Catalogue identity notes
Maker repo ai21labs/AI21-Jamba-Mini-1.6 @10e3208: licence tag other (jamba-open-model-license). 51,570,323,328 parameters from the Hugging Face safetensors metadata; active 12B per token (same architecture size as Jamba Mini (AI21 docs: '52B parameters (12B active)'; see catalogue jamba-mini-1-7)). Released 2025-02: month the maker's Hugging Face repo was created (createdAt 2025-02-27). Gated repository (licence acceptance required on Hugging Face).
$ how it compares
The LLMs ranked either side of Jamba 1.6 Mini, out of 1132 with a score, then more from AI21 Labs.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 483 |  DeepSeek-Coder-V2 Instruct (0724) | 53±1 | local | ||
| 484 |  Llama-PLLuM-70B-chat (2412) | 53±4 | local | ||
| 485 | Jamba 1.6 Mini | 53±10 | local | ||
| 486 |  Olmo 3 7B Think | 53±1 | local | ||
| 487 |  Functionary Small v3.1 | 53±6 | local | ||
| more from AI21 Labs | |||||
| 332 | Jamba 1.6 Large | 58±11 | local | ||
| 526 | Jamba2 Mini | 52±15 | api price unknown +local | ||
| 560 | Jamba 1.5 Large | 50±1 | local | ||
$ what it's good at
other tests1 not in a task · 52% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Vectara hallucination leaderboard, 2025 dataset (HHEM-2.3, factual consistency rate) | 95.4% ✎ | ▁▁▂▃▄▄█▆▃▂▂▁ beats 70/140 | 52% |
$ where to run it
llama.cpp · recommended
- Install llama.cpp · Homebrew and winget installs name the commands llama-cli and llama-server instead.macOS / Linux
curl -LsSf https://llama.app/install.sh | shWindowsirm https://llama.app/install.ps1 | iex - Run Jamba 1.6 Mini · needs ~31.2 GB · from huggingface.co/mradermacherchat in the terminal
llama cli -hf mradermacher/AI21-Jamba-Mini-1.6-GGUF:Q4_K_M - Or start a local server and open the address it prints (http://localhost:9931)
llama serve -hf mradermacher/AI21-Jamba-Mini-1.6-GGUF:Q4_K_M
LM Studio
- Install LM Studio · Open the app once before using the lms command.
- Run Jamba 1.6 Mini · needs ~31.2 GB · from huggingface.co/mradermacher
- Or download it in the terminal
lms get https://huggingface.co/mradermacher/AI21-Jamba-Mini-1.6-GGUF@Q4_K_M -y - Then run lms ls to see its name, and lms chat followed by that name
Ollama
- Install Ollama
download it (macOS / Windows) ↗
Linuxcurl -fsSL https://ollama.com/install.sh | sh - Run Jamba 1.6 Mini · needs ~31.2 GB · from huggingface.co/mradermacherchat in the terminal
ollama run hf.co/mradermacher/AI21-Jamba-Mini-1.6-GGUF:Q4_K_M
Other ways to run it
- Run it as published · needs ~103.14 GB · from huggingface.co/ai21labs (official)
uv run --with torch --with 'transformers<5' --with accelerate python -c "from transformers import pipeline; print(pipeline('text-generation', model='ai21labs/AI21-Jamba-Mini-1.6', torch_dtype='auto', device_map='auto')('Once upon a time', max_new_tokens=40)[0]['generated_text'])"
Official weights and documentation: huggingface.co/ai21labs ↗
$ how the score was worked out · Vectara hallucination leaderb… 52%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 Vectara hallucination leaderboard (vectara/hallucination-leaderboard repository) | 1 | 52% |
| prior · pull toward the middle | 48% | |
| weighted average | = 53.0 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json