Hunyuan Hy3 vs Hy4 preview
Side-by-side comparison of API pricing, specs, benchmarks, and capabilities
Hunyuan Hy3 is 79.4% cheaper on blended cost ($0.257 vs $1.25/Mtok) — note: different model classes (budget vs mid tier)
| Specification |
by Tencent
|
by Tencent
|
|---|---|---|
| Overview | ||
| Status | Current budget | Current mid tier Open weights |
| Released | Jul 7, 2026 | Aug 27, 2026 |
| Pricing per million tokens | ||
| Input | $0.147/Mtok | $0.834/Mtok |
| Output | $0.588/Mtok | $2.50/Mtok |
| Blended avg | $0.257/Mtok | $1.25/Mtok |
| Cached input | $0.037/Mtok | $0.042/Mtok |
| Price basis | USD is FX-derived from Tencent's RMB list price (¥1 / ¥0.25 / ¥4 per 1M) at 6.8031 CNY/USD (2026-07-07); re-checked on FX drift, not a promo. | Tencent lists this model in RMB on its TokenHub rate card — ¥6 in / ¥18 out / ¥0.3 cache-hit per 1M tokens. Those two region tabs are different rate cards in general (新加坡 reprices GLM-5.3 from ¥8/¥28/¥2 to ¥10.07524/¥31.66504/¥1.871116), but the Hy4 preview row is byte-identical on both, so this model has no separate international list. The USD above is that list at 7.197 CNY/USD, the rate implied to three figures by all three prices the FP8 endpoint Tencent itself operates quotes in USD ($0.834 / $2.501 / $0.042, the only endpoint serving this model). That is Tencent's own commercial rate, not a market snapshot: ECB spot on 2026-08-27 was 6.7203 CNY/USD, which would instead give $0.893 / $2.679 / $0.045. We publish what Tencent charges — FX_CURRENCY_POLICY rule (1) — not a spot re-conversion of its domestic card. |
| Specifications | ||
| Context window | 262K tokens | 1M tokens |
| Parameters | Proprietary | 770B (A49B) |
| Speed (TPS) | — | — |
| Modalities | ||
| Input |
text
|
text
|
| Benchmarks sources: Artificial Analysis, LMSYS Chatbot Arena (UC Berkeley), Tencent model card / LMSYS Chatbot Arena (UC Berkeley) | ||
| Independent composite scores measured for both models — each on its own scale | ||
| LMSYS Chatbot Arena | 1413 ELO | 1633.3 ELO |
| AA Intelligence Index | 42.2 | — |
| AA Agentic Index | 31.4 | — |
| AA-Omniscience Index (−100–100) | -18.5 | — |
| Per-benchmark results published for Hunyuan Hy3 only — no independent per-benchmark scores exist for Hy4 preview yet | ||
| GPQA Diamond | 87.2 vendor | — |
| SWE-Bench Verified | 74.4 vendor | — |
| Humanity's Last Exam | 30 vendor | — |
| Providers | ||
| Available from |
Tencent — $0.147/$0.588/Mtok
|
Tencent — $0.834/$2.50/Mtok
|
Cost at scale
1M tokens · 50/50 input/output| Volume | Hunyuan Hy3 | Hy4 preview | Savings |
|---|---|---|---|
| 1M tokens | $0.26 | $1.25 | $0.99 (79.2%) |
| 10M tokens | $2.57 | $12.51 | $9.94 (79.5%) |
| 100M tokens | $25.73 | $125.08 | $99.35 (79.4%) |
| 1000M tokens | $257.25 | $1250.75 | $993.5 (79.4%) |
When to pick which
distilled from the pricing and spec data above
Pick Hunyuan Hy3 if…
- cost dominates: $0.257/Mtok blended vs $1.25 — 79.4% less on the same 50/50 token mix
Pick Hy4 preview if…
- your workload re-reads context (agents, RAG, long chats): cached input costs $0.042/Mtok — 95% off its list input price
- you need the longer context: 1M tokens vs 262K (4×)
- you want open weights — self-host it, fine-tune it, or exit the API entirely; Hunyuan Hy3 is closed
Summary
Hunyuan Hy3 by Tencent costs $0.147/Mtok input and $0.588/Mtok output, with a 262K-token context window. It supports text input.
Hy4 preview by Tencent costs $0.834/Mtok input and $2.50/Mtok output, with a 1M-token context window. It supports text input.
On a blended cost basis, Hunyuan Hy3 is 79.4% cheaper than Hy4 preview.
Note: Pricing is per million tokens. Actual costs vary with usage patterns, prompt caching, and batch discounts. Always verify against official provider pricing pages.
More comparisons
pairs sharing a model with this page
GLM-4.7-Flash vs Ministral 3 3B
100% price gap
Solar Pro 4 vs Gemini 3.5 Flash-Lite
94% price gap
Solar Pro 4 vs DeepSeek V4 Flash
92% price gap
Claude Fable 5 vs DeepSeek V4 Pro
90% price gap
GPT-5.6 Terra vs GPT-5.6 Luna
90% price gap
Llama 3.3 70B vs Claude Sonnet 4.6
89% price gap
Llama 3.3 70B vs GPT-5.4
89% price gap
GPT-Realtime-2.1 mini vs GPT-Realtime-2.1
88% price gap