whatmodel.sh / Code Llama 34B Instruct

Code Llama 34B Instruct

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

Code Llama 34B Instruct is an open-weights language model from Meta, released August 2023. Its published size is 33.7B parameters and a 16k-token context window. Licence: llama2.

Catalogue identity notes

Arena entry 'codellama-34b-instruct' (text). LMArena's leaderboard table (arena-leaderboard space @6f2eeeb) labels 'codellama-34b-instruct' as 'CodeLlama-34B-instruct' (Meta, Llama 2 Community), linking https://huggingface.co/codellama/CodeLlama-34b-Instruct-hf. Maker repo codellama/CodeLlama-34b-Instruct-hf @d4c1c47: licence tag llama2. 33,743,970,304 parameters from the Hugging Face safetensors metadata (dense, so active = total). Released 2023-08: month the maker's Hugging Face repo was created (createdAt 2023-08-24).

39.9±0.7#532 of 719 LLMsfrom 16 benchmark results across 7 sources

$ how it compares

The LLMs ranked either side of Code Llama 34B Instruct, out of 828 with a score, then more from Meta.

#modelscorewhereRAMreleased
622Nous Hermes 2 Mixtral 8x7B DPO Nous Research40±2local~31GB2024-01
623Deita 7B v1.0 HKUST NLP40±2local~32GB2023-12
624Code Llama 34B Instruct Meta40±1local~22GB2023-08
625Code Llama 34B (base) Meta40±1local~23GB2023-08
626Starling-LM 7B beta Nexusflow40±1local~6GB2024-03
more from Meta
24Muse Spark 1.3 Meta77±1api $2.00/M2026-09
40Muse Spark 1.2 Meta75±1api $2.00/M2026-08
43Muse Spark 1.1 Meta75±1api $2.00/M2026-07

see all LLMs · find the ones that fit your machine

$ what it's good at

coding▁▁▁▃▆██▇▆▄▃▁ #362 of 458 · 59% of its score · 6 tests

Beats LFM2.5 8B-A1B, Qwen3-VL 2B Thinking, DeepSeek R1 Distill Qwen 7B here, though they rank higher overall. Trails Qwen2.5 1.5B Instruct, Phi-3-mini-128k-instruct (Apr 2024), OpenHermes 2.5 Mistral 7B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
HumanEval +151.8% ✎▁█▄▅▇█·▄▇▅▄▄ beats 23/42easy · avg 78%▇▇▇▇▇▇▇▇▇▇▇▇▇ 33%sharp · at its level · several runs
HumanEvalSynthesize (HumanEvalPack, 6-language average)41.3% ✎▁·▂▁▂▆▅█▄▄▁▂ beats 19/35medium · avg 64%▇▇▇ 7.7%at its level
MBPP lab57.0% ✎▃▃▃█··▃▅··▅▃ beats 8/11easy · avg 73%▇▇ 5.2%at its level
HumanEvalExplain (HumanEvalPack, 6-language average)30.3% ✎▄▇█▃▄▇▄▁▇▃·▁ beats 26/35medium · avg 49%▇▇ 4.8%at its level
HumanEvalFix (HumanEvalPack, 6-language average)28.2% ✎▁·▁▁▄██▄█▁▅▅ beats 23/35medium · avg 56%▇▇ 4.5%sharp · at its level
Tulu 2: CodexEval pass@1076.5% ✎▃▃▃·▆▃▃█▃▂▂▅ beats 20/25very easy · avg 86%▇ 3.2%sharp · at its level
chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #360 of 393 * · 8.3% of its score · 2 tests

Beats Gemma 4 E2B, Code Llama 70B Instruct here, though they rank higher overall. Trails Gemma 3 270M IT, SmolLM2 360M Instruct, Granite 4.0 350M here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Polyrating rating (bias-aware re-fit of Chatbot Arena votes, mid-2024)1070.0 ✎▂▁▂▁▃▇█▆▇▄▂▂ beats 33/95avg 1201▇▇ 4.2%blunt
LMArena (text)1136.8 ✎▁▁▁▃▃▃▃▆▆▇█▃ beats 29/379avg 1268▇▇ 4.1%blunt
math▁▂▃▅▆▅▄█▄▁▁▁ #336 of 398 * · 2.2% of its score · 1 test

Beats GPT-3.5 Turbo (0125), DBRX Instruct, Command R Plus (04-2024) here, though they rank higher overall. Trails Qwen3 0.6B, Gemma 7B (base), LFM2 700M here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
GSM8K38.0% ✎▆▂▅█▅▅▃▆···▂ beats 13/25easy · avg 73%▇ 2.2%sharp · at its level
other tests6 not in a task · 30% of its score
testresultvs the others in its rundifficultycountswhy
EvoEval (overall)33.3% ✎▂▂▃▄▄█▁▄▅▃▁▃ beats 21/42medium · avg 55%▇▇▇▇▇ 12%at its level
BIG-Bench Hard59.2% ✎▁·····▂█▃▅·▁ beats 17/25easy · avg 77%▇▇▇ 6.7%sharp · at its level
BigCodeBench (Complete)35.1% ✎▂▁▁▄▂▅▇█▆▂▂▂ beats 15/58medium · avg 50%▇▇ 4.6%at its level
BigCodeBench (Instruct)28.9% ✎▃▂·▆▆██▆▅▃▃▃ beats 5/33medium · avg 41%▇▇ 4.4%at its level
MMLU50.9% ✎▁▁▃▃▄█▄▁█··▁ beats 11/25easy · avg 74%▇ 2.4%at its level
Tulu 2: TyDiQA gold passage, 1-shot F155.1% ✎▂··▁▃▂▂▂█▅▁▂ beats 19/25medium · avg 61%▇ 0.8%blunt · at its level

$ where to run it

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 Code Llama 34B Instruct · needs ~20.2 GB · from huggingface.co/mradermacher
    chat in the terminal
    llama cli -hf mradermacher/CodeLlama-34b-Instruct-hf-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf mradermacher/CodeLlama-34b-Instruct-hf-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 Code Llama 34B Instruct · needs ~20.2 GB · from huggingface.co/mradermacher

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/mradermacher/CodeLlama-34b-Instruct-hf-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 Code Llama 34B Instruct · needs ~20.72 GB · from ollama.com
    chat in the terminal
    ollama run codellama:34b-instruct-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 Code Llama 34B Instruct · needs ~19.4 GB, Mac only · from huggingface.co/mlx-community
    chat in the terminal
    mlx_lm.chat --model mlx-community/CodeLlama-34b-Instruct-hf-4bit
  3. Or start an OpenAI-compatible server at http://localhost:8080/v1
    mlx_lm.server --model mlx-community/CodeLlama-34b-Instruct-hf-4bit

Weights and documentation: huggingface.co/codellama ↗

$ how the score was worked out · Top Leaderboard Ranking = Top… 33%, Granite Code Models: A Family… 17%, Code Llama: Open Foundation M… 17% 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
Top Leaderboard Ranking = Top Coding Proficiency, Always? EvoEval: Evolving Coding Benchmarks via LLM — authors' own runs2▇▇▇▇▇▇▇▇▇▇ 33%CC BY 4.0
Granite Code Models: A Family of Open Foundation Models for Code Intelligence — authors' own runs3▇▇▇▇▇ 17%CC BY 4.0
Code Llama: Open Foundation Models for Code — authors' own runs · lab-reported2▇▇▇▇▇ 17%CC BY 4.0
Tulu 2 — base and instruction-tuned evaluations5▇▇▇▇▇ 15%CC BY 4.0
BigCodeBench paper (Zhuo et al., ICLR 2025), arXiv 2406.15877v42▇▇▇ 9%CC BY 4.0
Polyrating: A Cost-Effective and Bias-Aware Rating System for LLM Evaluation (arXiv:2409.00696v3)1▇ 4%CC-BY-4.0
LMArena leaderboard dataset1▇ 4%CC BY 4.0
prior · pull toward the middle▇ 0%
weighted average= 39.9

$ 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