Claude Haiku 4.5 vs Tencent HY3
anthropic/claude-haiku-4.5 · tencent/hy3
As of November 1, 2026, Tencent HY3 is the cheaper API on UberLLM: $0.0012 per million input tokens and $0.0050 per million output, versus $0.0518 / $0.2592 for Claude Haiku 4.5, about 98% less when output is 3 tokens for every input token. These are the best prices on the market right now. Both work with one OpenAI-compatible key and UberLLM adds no fee. Choose Tencent HY3 to cut cost; choose Claude Haiku 4.5 when its capabilities fit your task better.
Claude Haiku 4.5 vs Tencent HY3, side by side
| Spec | Claude Haiku 4.5 | Tencent HY3 |
|---|---|---|
| Price · input / 1M | $0.0518 | $0.0012 |
| Price · output / 1M | $0.2592 | $0.0050 |
| Context length | 200,000 | 262,144 |
| Max output tokens | 64,000 | 131,072 |
| Input modalities | text, image | text |
| Output modalities | text | text |
| Tools / functions | Yes | Yes |
| Vision (image input) | Yes | — |
| Reasoning | Yes | Yes |
| % off list | 94.8% off | 99.1% off |
Best prices on UberLLM right now. No UberLLM fee on requests; adding credit by card costs 5.5% + $0.50. Updated November 1, 2026. A price more than 70% below list is what buyers were billed, not an error. Why so cheap?
Claude Haiku 4.5 vs Tencent HY3 — FAQ
Is Claude Haiku 4.5 or Tencent HY3 cheaper?
Tencent HY3 is cheaper on a blended token mix, by about 98%. A request costs what its provider charged, UberLLM adds no fee, and every request is listed with its cost in Activity.
What's the price difference between Claude Haiku 4.5 and Tencent HY3?
Claude Haiku 4.5: $0.0518 in / $0.2592 out per 1M tokens. Tencent HY3: $0.0012 in / $0.0050 out per 1M tokens. That's roughly a 98% blended difference. Prices are live; UberLLM adds no fee.
Can I use both Claude Haiku 4.5 and Tencent HY3 with one API?
Yes. Both are served through a single OpenAI-compatible endpoint — set the base URL to https://api.uberllm.dev/v1 and your UberLLM key, then switch between them by changing the model id. Anthropic SDKs work via https://api.uberllm.dev/v1/messages.