Model Comparison
GPT-5.4 mini (medium)
vs. Hy3-preview (Reasoning)
Comparing 2 AI models · 7 benchmarks · OpenAI, Tencent
Recommended Pick
Te Hy3-preview (Reasoning) 10 metric wins Strongest on: Value, Input price, Output price
Best Value
Hy3-preview (Reasoning)
100.0 value score
48.6 reasoning / $0.20/1M
Lowest Price
Hy3-preview (Reasoning)
$0.12/1M input price
Best Reasoning
Hy3-preview (Reasoning)
48.6 reasoning score
Blends available reasoning benchmarks
Best for Coding
GPT-5.4 mini (medium)
37.5 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
Hy3-preview (Reasoning) has the strongest quality-to-price mix at 100.0 out of 100 value points.
Price gap
Hy3-preview (Reasoning) is 6.1x cheaper on input tokens than GPT-5.4 mini (medium).
Speed gap
GPT-5.4 mini (medium) generates about 1.4x as many tokens per second as Hy3-preview (Reasoning).
Reasoning gap
Hy3-preview (Reasoning) leads GPT-5.4 mini (medium) by 5.5 points on reasoning.
Coding gap
GPT-5.4 mini (medium) leads Hy3-preview (Reasoning) by 1.0 points on coding.
Response Face-Off
Run one prompt through the selected models and compare response quality with live speed and cost context.
GPT-5.4 mini (medium)
OpenAI
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Hy3-preview (Reasoning)
Tencent
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | Op GPT-5.4 mini (medium) | Top Pick Te Hy3-preview (Reasoning) |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $0.75/1M | $0.12/1M |
| Output Cost | $4.50/1M | $0.43/1M |
| Blended (3:1) | $1.69/1M | $0.20/1M |
| Specifications | ||
| Organization | OpenAI | Tencent |
| Release Date | Mar 17, 2026 | Apr 23, 2026 |
| Performance & Speed | ||
| Throughput | 179.7 tok/s | 131.3 tok/s |
| TTFT | 12099ms | 2526ms |
| Latency | 12099ms | 17758ms |
| Composite Indices | ||
| Value Score | 10.5 | 100.0 |
| Reasoning Score | 43.1 | 48.6 |
| Intelligence | 29.8 | 33.6 |
| Coding | 37.5 | 36.5 |
| Standard Benchmarks | ||
| GPQA | 82.3% | 86.7% |
| HLE | 17.1% | 25.5% |
| SciCode | 44.2% | 41.2% |
| LCR | 61.3% | 54.7% |
| IFBench | 64.8% | 63.1% |
| TAU-bench v2 | 36.5% | 92.7% |
| TerminalBench Hard | 34.1% | 34.1% |
Key Takeaways
Hy3-preview (Reasoning) offers the best value at $0.12/1M, making it ideal for high-volume applications and cost-conscious projects.
Hy3-preview (Reasoning) has the strongest reasoning profile with a 48.6 reasoning score, combining the available reasoning-heavy benchmarks.
GPT-5.4 mini (medium) reaches a 37.5 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
GPT-5.4 mini (medium)
- Code generation
- Software development
Hy3-preview (Reasoning)
- Cost-sensitive applications
- High-volume processing
- Complex reasoning tasks
- Research & analysis
Recommended Comparisons
Popular alternatives
