Key Takeaways
Qwen2.5 VL 72B Instruct wins:
- Cheaper input tokens
- Cheaper output tokens
- Supports vision
Qwen3 Max wins:
- Larger context window
- Faster response time
- Higher intelligence benchmark
- Better at coding
- Better at math
Price Advantage
Qwen2.5 VL 72B Instruct
Benchmark Advantage
Qwen3 Max
Context Window
Qwen3 Max
Speed
Qwen3 Max
Pricing Comparison
Price Comparison
| Metric | Qwen2.5 VL 72B Instruct | Qwen3 Max | Winner |
|---|---|---|---|
| Input (per 1M tokens) | $0.15 | $1.20 | Qwen2.5 VL 72B Instruct |
| Output (per 1M tokens) | $0.60 | $6.00 | Qwen2.5 VL 72B Instruct |
| Cache Read (per 1M) | $75000.00 | $240000.00 | Qwen2.5 VL 72B Instruct |
Using a 3:1 input/output ratio, Qwen2.5 VL 72B Instruct is 89% cheaper overall.
Qwen2.5 VL 72B Instruct Providers
Chutes $0.15 (Cheapest)
Nebius $0.25
Parasail $0.80
Novita $0.80
Together $1.20
Qwen3 Max Providers
Alibaba $1.20 (Cheapest)
Benchmark Comparison
6
Benchmarks Compared
0
Qwen2.5 VL 72B Instruct Wins
0
Qwen3 Max Wins
Benchmark Scores
| Benchmark | Qwen2.5 VL 72B Instruct | Qwen3 Max | Winner |
|---|---|---|---|
Intelligence Index Overall intelligence score | - | 31.3 | - |
Coding Index Code generation & understanding | - | 26.4 | - |
Math Index Mathematical reasoning | - | 80.7 | - |
MMLU Pro Academic knowledge | - | 84.1 | - |
GPQA Graduate-level science | - | 76.4 | - |
LiveCodeBench Competitive programming | - | 76.7 | - |
Qwen3 Max significantly outperforms in coding benchmarks.
Cost vs Quality
X-axis:
Y-axis:
Loading chart...
Other models
Context & Performance
Context Window
Qwen2.5 VL 72B Instruct
32,768
tokens
Max output: 32,768 tokens
Qwen3 Max
262,144
tokens
Max output: 65,536 tokens
Qwen3 Max has a 88% larger context window.
Speed Performance
| Metric | Qwen2.5 VL 72B Instruct | Qwen3 Max | Winner |
|---|---|---|---|
| Tokens/second | N/A | 27.0 tok/s | |
| Time to First Token | N/A | 2.37s |
Capabilities
Feature Comparison
| Feature | Qwen2.5 VL 72B Instruct | Qwen3 Max |
|---|---|---|
| Vision (Image Input) | ||
| Tool/Function Calls | ||
| Reasoning Mode | ||
| Audio Input | ||
| Audio Output | ||
| PDF Input | ||
| Prompt Caching | ||
| Web Search |
License & Release
| Property | Qwen2.5 VL 72B Instruct | Qwen3 Max |
|---|---|---|
| License | Open Source | Open Source |
| Author | Qwen | Qwen |
| Released | Feb 2025 | Sep 2025 |
Qwen2.5 VL 72B Instruct Modalities
Input
textimage
Output
text
Qwen3 Max Modalities
Input
text
Output
text