whatmodel.sh / Aya Expanse 32B

Aya Expanse 32B

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Aya Expanse 32B is an open-weights language model from Cohere, released October 2024. Its published size is 32.3B parameters and a 128k-token context window. Licence: cc-by-nc-4.0.

Catalogue identity notes

Arena entry 'c4ai-aya-expanse-32b' (text). LMArena's leaderboard table (arena-leaderboard space @6f2eeeb) labels 'c4ai-aya-expanse-32b' as 'Aya-Expanse-32B' (Cohere, CC-BY-NC-4.0), linking https://huggingface.co/CohereForAI/aya-expanse-32b. Maker repo CohereLabs/aya-expanse-32b @b306ea2: licence tag cc-by-nc-4.0. 32,296,476,672 parameters from the Hugging Face safetensors metadata (dense, so active = total). Released 2024-10: month the maker's Hugging Face repo was created (createdAt 2024-10-23). Cohere models page lists c4ai-aya-expanse-32b as Live.

51.7±1.8#316 of 719 LLMsfrom 10 benchmark results across 7 sources

$ how it compares

The LLMs ranked either side of Aya Expanse 32B, out of 828 with a score, then more from Cohere.

#modelscorewhereRAMreleased
382Palmyra X4 (palmyra-x-004) Writer52±4comparison only2024-10
383Nemotron-4 340B Instruct NVIDIA52±2comparison only2024-06
384Aya Expanse 32B Cohere52±2api $0.75/M +local~22GB2024-10
385Gemma 3 12B Google52±1api $0.075/M +local~9.1GB2025-03
386Qwen2.5 Coder 7B Instruct Alibaba52±1local~6GB2024-09
more from Cohere
76North Small Translate 1.0 Cohere71±6api price unknown2026-08
175Command A+ (05-2026) Cohere64±1api price unknown +local~71GB2026-05
277Command A (03-2025) Cohere57±1api $4.38/M +local~37GB2025-03

see all LLMs · find the ones that fit your machine

$ what it's good at

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #254 of 393 * · 52% of its score · 2 tests

Beats Mistral Small 4 (119B-A6.5B), Ministral 3 14B Reasoning 2512, Ministral 3 14B here, though they rank higher overall. Trails Flan-T5 XXL, Granite 4.0 H 1B, Granite 4.0 1B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
LMArena (text)1267.2 ✎▁▁▁▂▃▃▃▆▆▇█▃ beats 102/379avg 1268▇▇▇▇▇▇▇▇▇▇▇ 28%blunt
Arabic IFEval prompt-level (Inception)56.7% ✎▁·▁▁▂▁▄▄▆▄█▄ beats 31/38medium · avg 49%▇▇▇▇▇▇▇▇▇▇ 24%sharp · at its level
other tests8 not in a task · 47% of its score
testresultvs the others in its rundifficultycountswhy
KCL-Essay vanilla20.2% ✎▁···▁▃▂█▂▃▃▂ beats 2/23hard · avg 18%▇▇▇▇▇▇ 14%typical
Open Persian LLM Leaderboard v2 average (21 tasks)62.9% ✎▂▄▂▃█▄▃▆▃▅▂▂ beats 59/81medium · avg 58%▇▇▇▇▇ 12%blunt · at its level
KCL-MCQA vanilla28.4% ✎▂···█▂▆▆██▆▂ beats 2/23very hard · avg 30%▇▇ 6.1%typical
Open PT-LLM Leaderboard average (9 Portuguese tasks)73.0% ✎▃▁▁▁▁▁▁▁▂█▅▁ beats 201/276medium · avg 73%▇▇ 5.8%at its level · noisy
EuroEval HellaSwag (English), test split, 5-shot generative MC, accuracy88.1% ✎▃·▁▁▂▁▂▅█▇▄▁ beats 88/100medium · avg 68%▇▇ 4.7%at its level · noisy
EuroEval Life in the UK (English knowledge MC), test split, 5-shot, accuracy86.7% ✎▁▁▁▁▁▁▁▂▄█▄▃ beats 104/146easy · avg 83%▇ 2.9%noisy
Vectara hallucination leaderboard, 2025 dataset (HHEM-2.3, factual consistency rate)91.5% ✎▁▁▂▃▄▄█▆▃▂▂▁ beats 31/140very easy · avg 94%▇ 1.6%blunt · maxed out · noisy
Vectara hallucination leaderboard (HHEM-2.3, factual consistency rate)89.1% ✎▁▂▁▂▅█▂▃▄▃▁▁ beats 32/98very easy · avg 90%▇ 0.0%blunt

$ where to run it

Through an API

api · $0.75/M · $0.5 in / $1.5 out per 1M tokens · 128k context · docs.cohere.com ↗ · price sourceUSD per 1M tokens, Cohere API: the pricing page FAQ states 'Aya Expanse models (8B and 32B) on the API are charged at $0.50/1M tokens for input and $1.50/1M tokens for output'. Cohere's models page lists c4ai-aya-expanse-32b on the Chat endpoint. Trial-key free usage excluded. Currency: USD. Cohere official pricing page, observed 2026-10-07.

llama.cpp · recommended

version

  1. 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 | sh
    Windows
    irm https://llama.app/install.ps1 | iex

    other ways to install ↗

  2. Run Aya Expanse 32B · needs ~19.8 GB · from huggingface.co/lmstudio-community
    chat in the terminal
    llama cli -hf lmstudio-community/aya-expanse-32b-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf lmstudio-community/aya-expanse-32b-GGUF:Q4_K_M

LM Studio

version

  1. Install LM Studio · Open the app once before using the lms command.

    download it (macOS / Windows / Linux) ↗

  2. Run Aya Expanse 32B · needs ~19.8 GB · from huggingface.co/lmstudio-community

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/lmstudio-community/aya-expanse-32b-GGUF@Q4_K_M -y
  4. Then run lms ls to see its name, and lms chat followed by that name

Ollama

version

  1. Install Ollama

    download it (macOS / Windows) ↗

    Linux
    curl -fsSL https://ollama.com/install.sh | sh
  2. Run Aya Expanse 32B · needs ~19.88 GB · from ollama.com
    chat in the terminal
    ollama run aya-expanse:32b-q4_K_M

MLX (Mac)

  1. Install MLX (Mac) · Apple silicon Macs only, text only.
    with uv
    uv tool install mlx-lm
    or
    pip install mlx-lm
  2. Run Aya Expanse 32B · needs ~34.3 GB, Mac only · from huggingface.co/mlx-community
    chat in the terminal
    mlx_lm.chat --model mlx-community/aya-expanse-32b-8bit
  3. Or start an OpenAI-compatible server at http://localhost:8080/v1
    mlx_lm.server --model mlx-community/aya-expanse-32b-8bit

Weights and documentation: huggingface.co/CohereLabs ↗

$ how the score was worked out · LMArena leaderboard dataset 28%, Inception Arabic/Hindi leader… 24%, Korean Canonical Legal Benchm… 20% and 4 more

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

sourceresultsshare of the scorelicence
LMArena leaderboard dataset1▇▇▇▇▇▇▇▇ 28%CC BY 4.0
Inception Arabic/Hindi leaderboards (AraGen-v3, HindiGen-v1, Arabic IFEval)1▇▇▇▇▇▇▇ 24%Apache-2.0
Korean Canonical Legal Benchmark: Toward Knowledge-Independent Evaluation of LLMs’ Legal Reasoning Capabilities2▇▇▇▇▇▇ 20%CC BY 4.0
Open Persian LLM Leaderboard v2 (Part DP AI and AUT NLP Lab)1▇▇▇▇ 12%Apache-2.0
EuroEval leaderboards (results collection), English generative leaderboard2▇▇ 8%MIT
Open Portuguese LLM Leaderboard (archived)1▇▇ 6%Apache-2.0
Vectara hallucination leaderboard (vectara/hallucination-leaderboard repository)2▇ 2%Apache-2.0
prior · pull toward the middle▇ 1%
weighted average= 51.7

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

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