Meta Llama 3 70B Instruct vs Qwen2 72B Instruct
Qwen2 72B Instruct leads llmrun Score 77.5 to 74.3 (ranges overlap); neither fits 24 GB.
Choose Qwen2 72B Instruct if you want…
Higher llmrun Score
77.5 vs 74.3, ranges overlap
Ahead on the benchmarks
3 of 3 shared boards
Meta Llama 3 70B Instruct: no clear edge over Qwen2 72B Instruct in the data we have.
On your hardware
Each model at its default quant (Q4_K_M), a smaller one when it must. Estimates, ±20%.
8 GB
RTX 4060 · 8 GB tier
Meta Llama 3 70B Instruct
Doesn't fit
smallest quant needs 28 GB
Qwen2 72B Instruct
Doesn't fit
smallest quant needs 25.7 GB
12 GB
RTX 3060 12GB · 12 GB tier
Meta Llama 3 70B Instruct
Doesn't fit
smallest quant needs 28 GB
Qwen2 72B Instruct
Doesn't fit
smallest quant needs 25.7 GB
16 GB
RTX 5060 Ti 16GB · 16 GB tier
Meta Llama 3 70B Instruct
Doesn't fit
smallest quant needs 28 GB
Qwen2 72B Instruct
Doesn't fit
smallest quant needs 25.7 GB
24 GBreference
RTX 4090 · 24 GB tier
Meta Llama 3 70B Instruct
Doesn't fit
smallest quant needs 28 GB
Qwen2 72B Instruct
Doesn't fit
smallest quant needs 25.7 GB
32 GB
RTX 5090 · 32 GB tier
Meta Llama 3 70B Instruct
IQ2_XS · ~50 tok/s
28 GB · up to 16K
Qwen2 72B Instruct
IQ2_M · ~46 tok/s
30.2 GB · up to 16K
48 GB
RTX A6000 · 48 GB tier
Meta Llama 3 70B Instruct
Q4_K_M · ~11 tok/stight
51.3 GB · up to 4K
Qwen2 72B Instruct
Q4_K_M · ~11 tok/stight
49.3 GB · up to 8K
Mac 32 GB
Mac Mini M4 · Macs
Meta Llama 3 70B Instruct
IQ2_XS · ~4 tok/s
28 GB · up to 16K
Qwen2 72B Instruct
IQ2_S · ~4 tok/s
28.4 GB · up to 16K
Mac 64 GB
Mac Studio M4 Max · Macs
Meta Llama 3 70B Instruct
Q4_K_M · ~8 tok/s
51.3 GB · up to 32K
Qwen2 72B Instruct
Q4_K_M · ~9 tok/s
49.3 GB · up to 32Kfaster
Mac 128 GB
Mac Studio M4 Max · Macs
Meta Llama 3 70B Instruct
Q4_K_M · ~8 tok/s
51.3 GB · up to 32K
Qwen2 72B Instruct
Q4_K_M · ~9 tok/s
49.3 GB · up to 32Kfaster
Reference GPU: RTX 4090. Tinted cell = the better pick on that machine (fits, more headroom, or at least 10% faster).
Overall
Meta Llama 3 70B Instruct
Qwen2 72B Instruct
Meta Llama 3 70B Instruct · #89 of 100 open models · #39 of 49 that fit ≤64 GB
Qwen2 72B Instruct · #88 of 100 open models · #38 of 49 that fit ≤64 GB
Benchmarks
3 shared boards: Meta Llama 3 70B Instruct leads 0, Qwen2 72B Instruct leads 3 · Benchmarks imported Oct 2026
Meta Llama 3 70B Instruct
40.6%#70/98 open
IndependentEpoch AI · Apr 2024
Qwen2 72B Instruct
40.8%#69/98 open
IndependentEpoch AI · Jun 2024
Meta Llama 3 70B Instruct
56.2%#64/145 open
IndependentTIGER-Lab
Qwen2 72B Instruct
64.4%#49/145 open
IndependentTIGER-Lab
Meta Llama 3 70B Instruct
22.6%#30/39 open
IndependentEpoch AI · Apr 2024
Qwen2 72B Instruct
39.1%#22/39 open
IndependentEpoch AI · Jun 2024
6 more boards (other shared and single-model boards)
Meta Llama 3 70B Instruct
1276.6#119/187 open
ArenaLMArena · Oct 2026
Qwen2 72B Instruct
1261.9#126/187 open
ArenaLMArena · Oct 2026
Meta Llama 3 70B Instruct
79.3%#13/85 open
Lab-reportedEpoch AI · Apr 2024
Qwen2 72B Instruct
82.4%#8/85 open
Lab-reportedEpoch AI · Jun 2024
Only Meta Llama 3 70B Instruct (2)
- AIME 2024/20254.3%
- DTBench54.2%
Only Qwen2 72B Instruct (2)
- WeirdML11.3%
- METR Time Horizons29.9%
VRAM by quantization
Memory to load the model and 2K context; below it, the total at 32K and 128K context. Lighter side in bold.
Q2_K
Meta Llama 3 70B Instruct
33.0 GB26.4 GB file
32K 43.1 GB · 128K 75.5 GB
Qwen2 72B Instruct
31.9 GB29.8 GB file
32K 41.9 GB · 128K —
Q3_K_S
Meta Llama 3 70B Instruct
34.0 GB30.9 GB file
32K 44.1 GB · 128K 76.4 GB
Qwen2 72B Instruct
32.8 GB34.5 GB file
32K 42.9 GB · 128K —
Q3_K_M
Meta Llama 3 70B Instruct
37.8 GB34.3 GB file
32K 48.0 GB · 128K 80.3 GB
Qwen2 72B Instruct
36.4 GB37.7 GB file
32K 46.5 GB · 128K —
Q4_0
Meta Llama 3 70B Instruct
38.8 GB40.0 GB file
32K 48.9 GB · 128K 81.3 GB
Qwen2 72B Instruct
37.3 GB41.4 GB file
32K 47.4 GB · 128K —
Q4_K_M
Meta Llama 3 70B Instruct
46.6 GB42.5 GB file
32K 56.7 GB · 128K 89.1 GB
Qwen2 72B Instruct
44.6 GB47.4 GB file
32K 54.7 GB · 128K —
Q5_K_M
Meta Llama 3 70B Instruct
55.3 GB50.0 GB file
32K 65.4 GB · 128K 97.8 GB
Qwen2 72B Instruct
52.8 GB54.4 GB file
32K 62.8 GB · 128K —
Q6_K
Meta Llama 3 70B Instruct
64.0 GB57.9 GB file
32K 74.2 GB · 128K 107 GB
Qwen2 72B Instruct
61.0 GB64.3 GB file
32K 71.0 GB · 128K —
Q8_0
Meta Llama 3 70B Instruct
77.6 GB75.0 GB file
32K 87.7 GB · 128K 120 GB
Qwen2 72B Instruct
73.7 GB77.3 GB file
32K 83.8 GB · 128K —
16 more quants
IQ2_XXS
Meta Llama 3 70B Instruct
not listed
Qwen2 72B Instruct
21.0 GB25.5 GB file
32K 31.0 GB · 128K —
IQ2_XS
Meta Llama 3 70B Instruct
23.3 GB21.1 GB file
32K 33.4 GB · 128K 65.8 GB
Qwen2 72B Instruct
22.8 GB27.1 GB file
32K 32.9 GB · 128K —
IQ2_S
Meta Llama 3 70B Instruct
not listed
Qwen2 72B Instruct
23.7 GB27.9 GB file
32K 33.8 GB · 128K —
IQ2_M
Meta Llama 3 70B Instruct
not listed
Qwen2 72B Instruct
25.5 GB29.3 GB file
32K 35.6 GB · 128K —
IQ3_XXS
Meta Llama 3 70B Instruct
not listed
Qwen2 72B Instruct
29.1 GB31.8 GB file
32K 39.2 GB · 128K —
IQ3_XS
Meta Llama 3 70B Instruct
not listed
Qwen2 72B Instruct
31.0 GB32.8 GB file
32K 41.0 GB · 128K —
IQ3_S
Meta Llama 3 70B Instruct
not listed
Qwen2 72B Instruct
31.9 GB34.5 GB file
32K 41.9 GB · 128K —
IQ3_M
Meta Llama 3 70B Instruct
not listed
Qwen2 72B Instruct
33.7 GB35.5 GB file
32K 43.8 GB · 128K —
Q3_K_L
Meta Llama 3 70B Instruct
39.8 GB37.1 GB file
32K 49.9 GB · 128K 82.3 GB
Qwen2 72B Instruct
38.2 GB39.5 GB file
32K 48.3 GB · 128K —
IQ4_XS
Meta Llama 3 70B Instruct
not listed
Qwen2 72B Instruct
40.1 GB39.7 GB file
32K 50.1 GB · 128K —
Q4_1
Meta Llama 3 70B Instruct
43.7 GB44.3 GB file
32K 53.8 GB · 128K 86.2 GB
Qwen2 72B Instruct
not listed
Q4_K_S
Meta Llama 3 70B Instruct
43.7 GB40.3 GB file
32K 53.8 GB · 128K 86.2 GB
Qwen2 72B Instruct
41.9 GB43.9 GB file
32K 51.9 GB · 128K —
Q5_0
Meta Llama 3 70B Instruct
48.5 GB48.7 GB file
32K 58.6 GB · 128K 91.0 GB
Qwen2 72B Instruct
46.4 GB50.2 GB file
32K 56.5 GB · 128K —
Q5_1
Meta Llama 3 70B Instruct
53.4 GB53.0 GB file
32K 63.5 GB · 128K 95.9 GB
Qwen2 72B Instruct
not listed
Q5_K_S
Meta Llama 3 70B Instruct
53.4 GB48.7 GB file
32K 63.5 GB · 128K 95.9 GB
Qwen2 72B Instruct
51.0 GB51.4 GB file
32K 61.0 GB · 128K —
BF16
Meta Llama 3 70B Instruct
155 GB~141.1 GB file
32K 165 GB · 128K 198 GB
Qwen2 72B Instruct
146 GB~145.4 GB file
32K 156 GB · 128K —
Meta Llama 3 70B Instruct
Ollama · Q4_K_M
ollama run hf.co/second-state/Meta-Llama-3-70B-Instruct-GGUF:Q4_K_MQwen2 72B Instruct
Ollama · Q4_K_M
ollama run hf.co/mradermacher/Qwen2-72B-Instruct-i1-GGUF:Q4_K_M| Spec | Meta Llama 3 70B Instruct | Qwen2 72B Instruct |
|---|---|---|
| Publisher | Meta | Alibaba |
| Parameters | 70.6B | 72.7B |
| Architecture | — | Qwen2ForCausalLM |
| Context | — | 32K tokens |
| Modalities | Chat | Chat |
| License | Llama 3 Community | Other |
| Released | Apr 2024 | May 2024 |
Frequently Asked Questions
Which is better, Meta Llama 3 70B Instruct or Qwen2 72B Instruct?
Qwen2 72B Instruct leads llmrun Score 77.5 to 74.3 (ranges overlap); neither fits 24 GB. Choose Qwen2 72B Instruct for higher llmrun Score and ahead on the benchmarks.
Which needs less VRAM, Meta Llama 3 70B Instruct or Qwen2 72B Instruct?
At Q4_K_M, Meta Llama 3 70B Instruct requires 46.6 GB while Qwen2 72B Instruct needs 44.6 GB, making Qwen2 72B Instruct the easier model to run on your local setup.
Can I run Meta Llama 3 70B Instruct or Qwen2 72B Instruct on a 16 GB GPU?
Meta Llama 3 70B Instruct does not fit (smallest quant needs 28 GB); Qwen2 72B Instruct does not fit (smallest quant needs 25.7 GB).