Model Comparison
Hy3-preview (Reasoning)
vs. Tiny Aya Global
Comparing 2 AI models · 7 benchmarks · Tencent, Cohere
Recommended Pick
Te Hy3-preview (Reasoning) 10 metric wins Strongest on: Reasoning, Intelligence, Coding
Lowest Price
Tiny Aya Global
$0.00/1M input price
Best Reasoning
Hy3-preview (Reasoning)
51.4 reasoning score
Blends available reasoning benchmarks
Best for Coding
Hy3-preview (Reasoning)
36.5 coding index
Composite Indices
Higher is better; speed and price are normalized
Standard Benchmarks
Only benchmarks with data are shown
Differences That Matter
Price gap
Tiny Aya Global is ∞x cheaper on input tokens than Hy3-preview (Reasoning).
Speed gap
Tiny Aya Global generates about 1.3x as many tokens per second as Hy3-preview (Reasoning).
Reasoning gap
Hy3-preview (Reasoning) leads Tiny Aya Global by 37.9 points on reasoning.
Coding gap
Hy3-preview (Reasoning) leads Tiny Aya Global by 35.3 points on coding.
Top-pick rationale
Hy3-preview (Reasoning) wins 10 measurable categories, including Reasoning, Intelligence, Coding, GPQA.
Response Face-Off
Run one prompt through the selected models and compare response quality with live speed and cost context.
Hy3-preview (Reasoning)
Tencent
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Tiny Aya Global
Cohere
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | Top Pick Te Hy3-preview (Reasoning) | Co Tiny Aya Global |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $0.12/1M | $0.00/1M |
| Output Cost | $0.43/1M | $0.00/1M |
| Blended (3:1) | $0.20/1M | — |
| Specifications | ||
| Organization | Tencent | Cohere |
| Release Date | Apr 23, 2026 | Feb 17, 2026 |
| Performance & Speed | ||
| Throughput | 93.2 tok/s | 125.3 tok/s |
| TTFT | 2483ms | 329ms |
| Latency | 23953ms | 329ms |
| Composite Indices | ||
| Value Score | 100.0 | — |
| Reasoning Score | 51.4 | 13.5 |
| Intelligence | 41.9 | 4.7 |
| Coding | 36.5 | 1.2 |
| Standard Benchmarks | ||
| GPQA | 86.7% | 30.5% |
| HLE | 25.5% | 5.2% |
| SciCode | 41.2% | 3.6% |
| LCR | 54.7% | 0.0% |
| IFBench | 63.1% | 20.1% |
| TAU-bench v2 | 92.7% | 0.0% |
| TerminalBench Hard | 34.1% | 0.0% |
Key Takeaways
Tiny Aya Global offers the best value at $0.00/1M, making it ideal for high-volume applications and cost-conscious projects.
Hy3-preview (Reasoning) has the strongest reasoning profile with a 51.4 reasoning score, combining the available reasoning-heavy benchmarks.
Hy3-preview (Reasoning) reaches a 36.5 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
Hy3-preview (Reasoning)
- Complex reasoning tasks
- Research & analysis
- Code generation
- Software development
Tiny Aya Global
- Cost-sensitive applications
- High-volume processing
