Model Comparison
Qwen3 VL 235B A22B Instruct
vs. Trinity Large Thinking
Comparing 2 AI models · 10 benchmarks · Alibaba, Arcee AI
Recommended Pick
Ar Trinity Large Thinking 15 metric wins Strongest on: Value, Input price, Output price
Best Value
Trinity Large Thinking
100.0 value score
40.6 reasoning / $0.40/1M
Lowest Price
Trinity Large Thinking
$0.23/1M input price
Best Reasoning
Qwen3 VL 235B A22B Instruct
47.9 reasoning score
Blends available reasoning benchmarks
Best for Coding
Trinity Large Thinking
27.2 coding index
Composite Indices
Higher is better; speed and price are normalized
Standard Benchmarks
Only benchmarks with data are shown
Differences That Matter
Best value
Trinity Large Thinking has the strongest quality-to-price mix at 100.0 out of 100 value points.
Price gap
Trinity Large Thinking is 1.3x cheaper on input tokens than Qwen3 VL 235B A22B Instruct.
Speed gap
Trinity Large Thinking generates about 3.6x as many tokens per second as Qwen3 VL 235B A22B Instruct.
Reasoning gap
Qwen3 VL 235B A22B Instruct leads Trinity Large Thinking by 7.3 points on reasoning.
Coding gap
Trinity Large Thinking leads Qwen3 VL 235B A22B Instruct by 10.7 points on coding.
Response Face-Off
Run one prompt through the selected models and compare response quality with live speed and cost context.
Qwen3 VL 235B A22B Instruct
Alibaba
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Trinity Large Thinking
Arcee AI
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | Al Qwen3 VL 235B A22B Instruct | Top Pick Ar Trinity Large Thinking |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $0.30/1M | $0.23/1M |
| Output Cost | $1.90/1M | $0.88/1M |
| Blended (3:1) | $0.70/1M | $0.40/1M |
| Specifications | ||
| Organization | Alibaba | Arcee AI |
| Release Date | Sep 23, 2025 | Apr 1, 2026 |
| Performance & Speed | ||
| Throughput | 53.5 tok/s | 191.4 tok/s |
| TTFT | 1090ms | 789ms |
| Latency | 1090ms | 11240ms |
| Composite Indices | ||
| Value Score | 66.6 | 100.0 |
| Reasoning Score | 47.9 | 40.6 |
| Intelligence | 20.8 | 31.9 |
| Coding | 16.5 | 27.2 |
| Math | 70.7 | — |
| Standard Benchmarks | ||
| GPQA | 71.2% | 75.2% |
| MMLU Pro | 82.3% | — |
| HLE | 6.3% | 14.7% |
| LiveCodeBench | 59.4% | — |
| AIME 2025 | 70.7% | — |
| SciCode | 35.9% | 36.1% |
| LCR | 31.7% | 33.0% |
| IFBench | 42.7% | 56.3% |
| TAU-bench v2 | 35.1% | 90.1% |
| TerminalBench Hard | 6.8% | 22.7% |
Key Takeaways
Trinity Large Thinking offers the best value at $0.23/1M, making it ideal for high-volume applications and cost-conscious projects.
Qwen3 VL 235B A22B Instruct has the strongest reasoning profile with a 47.9 reasoning score, combining the available reasoning-heavy benchmarks.
Trinity Large Thinking reaches a 27.2 coding index, making it the top choice for software development and code generation tasks.
All models support context windows of ∞+ tokens, suitable for processing lengthy documents and maintaining extended conversations.
When to Choose Each Model
Qwen3 VL 235B A22B Instruct
- Complex reasoning tasks
- Research & analysis
Trinity Large Thinking
- Cost-sensitive applications
- High-volume processing
- Code generation
- Software development
