- Size
- 72.7B parameterscounted from the safetensors header on the Hub
- Licence
- custom, read it, Tongyi Qianwen License AgreementOur reading, check the licence: A custom licence we have not read for this model: other makers use the same name for different texts. Read the model's own licence file before you use it.Licence text
- On the Hub since
- 28 May 2024the date the repo was created
- Context length
- 32k tokensfrom the model's config
- Fine-tune
- of Qwen/Qwen2-72BQwen2ForCausalLM
- On Hugging Face
- Qwen/Qwen2-72B-Instruct24k downloads in the last 30 days, 724 likes
Scores
- default setting40.8%±2.627 Jan 2025, log
Epoch's name: qwen2-72b-instruct. Best closed model on this set: gpt-6.2-astra_max, 96.8% (OpenAI). All 341 results
- default setting39.1%±1.027 Jan 2025, log
Epoch's name: qwen2-72b-instruct. Best closed model on this set: gpt-5-2025-08-07_high, 98.1% (OpenAI). All 108 results
- default setting29.9%METR - Measuring AI Ability to Complete Long Tasks
Epoch's name: qwen2-72b-instruct. Best closed model on this set: claude-mythos-preview-early, 85.2% (Anthropic). All 49 results
- Shots: 582.4%Stanford CRFM Leaderboard
Epoch's name: qwen2-72b-instruct. Best closed model on this set: gpt-4o-2024-11-20, 88.1% (OpenAI). All 217 results
- Environment model: Claude 3.5 Sonnet1.1%17 Dec 2024, TheAgentCompany experiment results github
Epoch's name: qwen2-72b-instruct. Best closed model on this set: gemini-2.5-flash-preview-09-2025, 41.1% (Google DeepMind). All 16 results
- default setting11.3%±0.0htihle.github.io
Epoch's name: qwen2-72b-instruct. Best closed model on this set: gpt-6-astra_promax, 93.6% (OpenAI). All 173 results
Claimed by the maker Claimed
The model card has no claimed scores in a machine-readable form (the card's model-index), so none are shown. The maker may give scores in the card text: read the card.
Other views, which we link to and do not copy: look this model up on Artificial Analysis and the Arena leaderboard.
Independent scores from Epoch AI (CC-BY 4.0), fetched 19 Nov 2026: Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Each row links its own source. What the badges mean.
Run it yourself
Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 2.5 GB, worked out from the model's config (Qwen/Qwen2-72B-Instruct), at 16-bit precision. Choose a machine above to see a verdict for each format. Method
GGUF for llama.cpp, Ollama, LM Studio and others
legraphista/Qwen2-72B-Instruct-IMat-GGUF (converted by legraphista), 1.2k downloads. Sizes are the real files on the Hub, all parts added up.
| Format | Size | Memory needed | Get it |
|---|---|---|---|
| IQ3_XXS | 29.7 GBfile size | 32.7 GB29.7 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ3_XXSOllama ollama run hf.co/legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ3_XXSDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.IQ3_XXS.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| IQ3_XS | 30.6 GBfile size | 33.6 GB30.6 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ3_XSOllama ollama run hf.co/legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ3_XSDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.IQ3_XS.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| IQ3_S | 32.1 GBfile size | 35.1 GB32.1 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ3_SOllama ollama run hf.co/legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ3_SDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.IQ3_S.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| Q3_K_S | 32.1 GBfile size | 35.1 GB32.1 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q3_K_SOllama ollama run hf.co/legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q3_K_SDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.Q3_K_S.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| IQ3_M | 33.1 GBfile size | 36.1 GB33.1 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ3_MOllama ollama run hf.co/legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ3_MDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.IQ3_M.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| Q3_K | 35.1 GBfile size | 38.1 GB35.1 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q3_KOllama ollama run hf.co/legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q3_KDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.Q3_K.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| Q3_K_L | 36.8 GBfile size | 39.8 GB36.8 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q3_K_LOllama ollama run hf.co/legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q3_K_LDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.Q3_K_L.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| IQ4_XS | 37.0 GBfile size | 40.0 GB37.0 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ4_XSOllama ollama run hf.co/legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ4_XSDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.IQ4_XS.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| IQ4_NL | 38.5 GBfile size | 41.5 GB38.5 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ4_NLOllama ollama run hf.co/legraphista/Qwen2-72B-Instruct-IMat-GGUF:IQ4_NLDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.IQ4_NL.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| Q4_K_S | 40.9 GBfile size | 43.9 GB40.9 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q4_K_SOllama ollama run hf.co/legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q4_K_SDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.Q4_K_S.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| Q4_K | 44.2 GBfile size, 2 parts | 47.2 GB44.2 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q4_KDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.Q4_K/Qwen2-72B-Instruct.Q4_K-*-of-00002.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| Q5_K_S | 47.8 GBfile size, 2 parts | 50.8 GB47.8 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q5_K_SDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.Q5_K_S/Qwen2-72B-Instruct.Q5_K_S-*-of-00002.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| Q5_K | 50.7 GBfile size, 3 parts | 53.7 GB50.7 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q5_KDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.Q5_K/Qwen2-72B-Instruct.Q5_K-*-of-00003.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| Q6_K | 59.9 GBfile size, 3 parts | 62.9 GB59.9 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q6_KDownload hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.Q6_K/Qwen2-72B-Instruct.Q6_K-*-of-00003.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| Q8_0 | 72.0 GBfile size, 4 parts | 75.0 GB72.0 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:Q8_0Download hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.Q8_0/Qwen2-72B-Instruct.Q8_0-*-of-00004.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| BF16 | 135.4 GBfile size, 6 parts | 138.4 GB135.4 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:BF16Download hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.BF16/Qwen2-72B-Instruct.BF16-*-of-00006.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
| F16 | 135.4 GBfile size, 6 parts | 138.4 GB135.4 + 2.5 + 0.50 | Commandsllama.cpp llama-server -hf legraphista/Qwen2-72B-Instruct-IMat-GGUF:F16Download hf download legraphista/Qwen2-72B-Instruct-IMat-GGUF --include "Qwen2-72B-Instruct.FP16/Qwen2-72B-Instruct.FP16-*-of-00006.gguf" --local-dir Qwen2-72B-Instruct-IMat-GGUF |
Original weights size from the safetensors header, for vLLM, SGLang, Transformers
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| BF16original weights | 135.4 GBfile size | 138.4 GB135.4 + 2.5 + 0.50 |
hf download Qwen/Qwen2-72B-Instruct --local-dir Qwen2-72B-InstructOther conversions (7; sizes not read)
- Qwen/Qwen2-72B-Instruct-GPTQ-Int4 GPTQ, int4, by the maker, 322 downloads
- Qwen/Qwen2-72B-Instruct-AWQ AWQ, by the maker, 309 downloads
- Qwen/Qwen2-72B-Instruct-GPTQ-Int8 GPTQ, int8, by the maker, 221 downloads
- mlc-ai/Qwen2-72B-Instruct-q4f16_1-MLC Other, 20 downloads
- mlc-ai/Qwen2-72B-Instruct-q0f16-MLC Other, 12 downloads
- Enfint/Qwen2-72B-Instruct-GPTQ-Int4 GPTQ, int4, 11 downloads
- mlc-ai/Qwen2-72B-Instruct-q4f32_1-MLC Other, 9 downloads
The Hub tags 24 quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.
Run it by API
No provider on Hugging Face Inference Providers serves this model (as of 19 Nov 2026, 13:58 UTC). Other hosts may; we only list what the router reports.
Not built yet: one key for every open model, with BlockRun, and our own tested-quality score per provider. Until then, precision and measured speed are the best guide to a provider here.
Planned for this page Not built yet
- Try it in private chat
- A link to Solana AI's confidential chat for the models it serves. This model is not listed as served there.
- Fine-tune it
- Fine-tuning a few base models from this page and publishing the result here.
- Creators and earnings
- Pages for people who publish fine-tunes and adapters of this model, and what they earn.
- Fund and bounties
- Bounties for new open models, paid from a public fund. How the fund will work.
- Contribute compute
- How to give compute to open model work, with Akash.
Model data from the Hugging Face Hub 19 Nov 2026, 13:43 UTC. The repo. Found a mistake on this page? Tell us.