Benchmark scores
Higher is better for all benchmark scores (SWE-bench, GPQA, HLE, Terminal-Bench, etc.). Green marks the model with the higher score.
LLM Comparison
GPT-5.5 xHigh vs Hy3 Preview: benchmark scores, pricing & comparison.
Side-by-side GPT-5.5 xHigh vs Hy3 Preview comparison across SWE-bench, GPQA, HLE, Terminal-Bench, coding agent scores, token pricing, context window, and AskClash RWT. Green marks the winner on each benchmark.
Rank #3 vs #30AskClash overall scores 86.0 vs 40.3.
Pricing $5.00/$30.0 vs $0/$0Input and output token prices per 1M tokens when published.
Proprietary vs Open WeightOpenAI vs Tencent.
GPT-5.5 xHigh vs Hy3 Preview benchmark comparison
Green cells highlight the winning model for each metric. Scores are cached from the AskClash LLM leaderboard snapshot.
| Metric | GPT-5.5 xHigh | Hy3 Preview |
|---|---|---|
| Overall Score | 86.0 | 40.3 |
| Leaderboard Rank | #3 | #30 |
| RWT | 9.0 | — |
| Coding Agent Index | 76.4 | — |
| HLE | 52.2 | 25.5 |
| GPQA | 93.6 | 87.2 |
| IFEval | — | 63.1 |
| SWE-bench | — | 74.4 |
| SWE-Pro | 58.6 | — |
| Terminal-Bench | 82.7 | 54.4 |
| OSWorld | 78.7 | — |
| MCP Atlas | 75.3 | — |
| Finance Agent | 51.8 | — |
| MMMU-Pro | 81.2 | — |
| ARC-AGI 2 | 85.0 | — |
| Tau2 | 98.0 | 92.7 |
| Input Price (per 1M tokens) | $5.00 | $0 |
| Output Price (per 1M tokens) | $30.0 | $0 |
| Context Window | 1M | 256K |
| Benchmark Cells | 12 | 6 |
More GPT-5.5 xHigh and Hy3 Preview comparisons
Explore how GPT-5.5 xHigh and Hy3 Preview stack up against other top-ranked LLMs.
Claude Mythos/Fable 5 vs GPT-5.5 xHighClaude Mythos/Fable 5 vs Hy3 PreviewClaude Opus 4.8 (Adaptive) vs GPT-5.5 xHighClaude Opus 4.8 (Adaptive) vs Hy3 PreviewGPT-5.5 xHigh vs GLM-5.2GLM-5.2 vs Hy3 PreviewGPT-5.5 xHigh vs Claude Opus 4.7 (Adaptive)Claude Opus 4.7 (Adaptive) vs Hy3 PreviewGPT-5.5 xHigh vs Qwen3.7 MaxQwen3.7 Max vs Hy3 Preview
How to read this comparison
Token pricing
Lower is better for input and output prices. Green marks the cheaper model per 1M tokens.
Coverage matters
Models with fewer disclosed benchmark cells may have inflated percentile scores. Check the benchmark cell count for context.
This comparison page is generated from the AskClash LLM leaderboard cache. Open the live leaderboard for real-time scores and interactive filtering.