Compare · ModelsLive · 2 picked · head to head
R1 vs Qwen3 235B A22B Instruct 2507
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Qwen3 235B A22B Instruct 2507 wins on 3/6 benchmarks
Qwen3 235B A22B Instruct 2507 wins 3 of 6 shared benchmarks. Leads in coding · arena.
Category leads
coding·Qwen3 235B A22B Instruct 2507reasoning·R1arena·Qwen3 235B A22B Instruct 2507knowledge·R1
Hype vs Reality
Attention vs performance
R1
#166 by perf·#16 by attention
Qwen3 235B A22B Instruct 2507
#147 by perf·#1 by attention
Best value
Qwen3 235B A22B Instruct 2507
3.6x better value than R1
R1
28.5 pts/$
$1.60/M
Qwen3 235B A22B Instruct 2507
102.8 pts/$
$0.47/M
Vendor risk
Mixed exposure
One or more vendors flagged
DeepSeek
$3.4B·Tier 1
Alibaba (Qwen)
$293.0B·Tier 1
Head to head
6 benchmarks · 2 models
R1Qwen3 235B A22B Instruct 2507
Aider polyglot
Qwen3 235B A22B Instruct 2507 leads by +2.7
Aider Polyglot · measures how well AI models can edit code across multiple programming languages using the Aider coding assistant framework.
R1
56.9
Qwen3 235B A22B Instruct 2507
59.6
ARC-AGI
R1 leads by +4.8
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
R1
15.8
Qwen3 235B A22B Instruct 2507
11.0
ARC-AGI-2
R1 leads by +0.1
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
R1
1.3
Qwen3 235B A22B Instruct 2507
1.3
Chatbot Arena Elo · Overall
Qwen3 235B A22B Instruct 2507 leads by +24.3
R1
1398.1
Qwen3 235B A22B Instruct 2507
1422.4
Fiction.LiveBench
R1 leads by +16.5
Fiction.LiveBench · a continuously updated benchmark using recently published fiction to test reading comprehension and reasoning, preventing data contamination.
R1
69.4
Qwen3 235B A22B Instruct 2507
52.9
WeirdML
Qwen3 235B A22B Instruct 2507 leads by +2.2
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
R1
36.5
Qwen3 235B A22B Instruct 2507
38.7
Full benchmark table
| Benchmark | R1 | Qwen3 235B A22B Instruct 2507 |
|---|---|---|
Aider polyglot Aider Polyglot · measures how well AI models can edit code across multiple programming languages using the Aider coding assistant framework. | 56.9 | 59.6 |
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 15.8 | 11.0 |
ARC-AGI-2 ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data. | 1.3 | 1.3 |
Chatbot Arena Elo · Overall | 1398.1 | 1422.4 |
Fiction.LiveBench Fiction.LiveBench · a continuously updated benchmark using recently published fiction to test reading comprehension and reasoning, preventing data contamination. | 69.4 | 52.9 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 36.5 | 38.7 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.70 | $2.50 | 64K tokens (~32 books) | $11.50 | |
| $0.14 | $0.80 | 262K tokens (~131 books) | $3.05 |