Compare · ModelsLive · 3 picked · head to head
Qwen2.5 72B Instruct Abliterated vs Qwen2.5 72B Instruct vs Qwen2.5 32B Instruct
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Qwen2.5 72B Instruct wins on 5/9 benchmarks
Qwen2.5 72B Instruct wins 5 of 9 shared benchmarks. Leads in general · language · math.
Category leads
general·Qwen2.5 72B Instructknowledge·Qwen2.5 72B Instruct Abliteratedlanguage·Qwen2.5 72B Instructmath·Qwen2.5 72B Instructreasoning·Qwen2.5 32B Instruct
Hype vs Reality
Attention vs performance
Qwen2.5 72B Instruct Abliterated
#139 by perf·no signal
Qwen2.5 72B Instruct
#128 by perf·#2 by attention
Qwen2.5 32B Instruct
#222 by perf·#2 by attention
Best value
Qwen2.5 72B Instruct
Qwen2.5 72B Instruct Abliterated
n/a
no price
Qwen2.5 72B Instruct
129.5 pts/$
$0.38/M
Qwen2.5 32B Instruct
n/a
no price
Vendor risk
Who is behind the model
HA
HuiHui AI
private · undisclosed
Alibaba (Qwen)
$293.0B·Tier 1
Alibaba (Qwen)
$293.0B·Tier 1
Head to head
9 benchmarks · 3 models
Qwen2.5 72B Instruct AbliteratedQwen2.5 72B InstructQwen2.5 32B Instruct
BBH (HuggingFace)
Qwen2.5 72B Instruct leads by +1.4
Qwen2.5 72B Instruct Abliterated
60.5
Qwen2.5 72B Instruct
61.9
Qwen2.5 32B Instruct
56.5
GPQA
Qwen2.5 72B Instruct Abliterated leads by +2.7
Qwen2.5 72B Instruct Abliterated
19.4
Qwen2.5 72B Instruct
16.7
Qwen2.5 32B Instruct
11.7
IFEval
Qwen2.5 72B Instruct leads by +0.4
Qwen2.5 72B Instruct Abliterated
85.9
Qwen2.5 72B Instruct
86.4
Qwen2.5 32B Instruct
83.5
MATH Level 5
Qwen2.5 32B Instruct leads by +2.4
Qwen2.5 72B Instruct Abliterated
60.1
Qwen2.5 72B Instruct
59.8
Qwen2.5 32B Instruct
62.5
MMLU-PRO
Qwen2.5 32B Instruct leads by +0.5
Qwen2.5 72B Instruct Abliterated
50.4
Qwen2.5 72B Instruct
51.4
Qwen2.5 32B Instruct
51.9
MUSR
Qwen2.5 32B Instruct leads by +1.2
Qwen2.5 72B Instruct Abliterated
12.3
Qwen2.5 72B Instruct
11.7
Qwen2.5 32B Instruct
13.5
GPQA diamond
Qwen2.5 72B Instruct leads by +4.1
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Qwen2.5 72B Instruct
32.2
Qwen2.5 32B Instruct
28.1
MATH level 5
Qwen2.5 72B Instruct leads by +7.1
MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics.
Qwen2.5 72B Instruct
63.2
Qwen2.5 32B Instruct
56.1
OTIS Mock AIME 2024-2025
Qwen2.5 72B Instruct leads by +0.7
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Qwen2.5 72B Instruct
8.0
Qwen2.5 32B Instruct
7.3
Full benchmark table
| Benchmark | Qwen2.5 72B Instruct Abliterated | Qwen2.5 72B Instruct | Qwen2.5 32B Instruct |
|---|---|---|---|
BBH (HuggingFace) | 60.5 | 61.9 | 56.5 |
GPQA | 19.4 | 16.7 | 11.7 |
IFEval | 85.9 | 86.4 | 83.5 |
MATH Level 5 | 60.1 | 59.8 | 62.5 |
MMLU-PRO | 50.4 | 51.4 | 51.9 |
MUSR | 12.3 | 11.7 | 13.5 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | — | 32.2 | 28.1 |
MATH level 5 MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics. | — | 63.2 | 56.1 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | — | 8.0 | 7.3 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
HA Qwen2.5 72B Instruct Abliterated | — | — | — | — |
| $0.36 | $0.40 | 33K tokens (~16 books) | $3.70 | |
| — | — | — | — |