All models
Tencent
Hy3 fastruns locally
tencent/hy3
Hy3 is a 295B-parameter Mixture-of-Experts model from Tencent (21B active, 192 experts with top-8 routing) built for reasoning, agentic workflows, and real-world production use. It supports a configurable reasoning effort:...
Input / 1M
$0.13
USD per million tokens
Output / 1M
$0.53
USD per million tokens
Est. cost / token
$0.000000231
$0.23 per 1M, 3:1 mix
Context window
262K
max output 128K
Major features
Open weightsCachingReasoningJSON schemaTools
Task suitability (heuristic 1–5, not a benchmark)
- coding3
- reasoning3
- math3
- summarization4
- extraction4
- classification4
- chat3
- creative3
- translation4
- vision1
- long context3
- agentic3
Run it locally
Will it run on my machine? →| Quantization | Quality | Memory needed (8K ctx) |
|---|---|---|
| Q2_K | poor | 137 GB |
| Q3_K_M | fair | 159 GB |
| Q4_K_Mrecommended | good | 197 GB |
| Q5_K_M | very good | 229 GB |
| Q6_K | excellent | 263 GB |
| Q8_0 | near-lossless | 339 GB |
| F16 | lossless | 633 GB |
Estimates.
Price history
| Recorded | Input | Output | Cached | Source |
|---|---|---|---|---|
| 2026-11-05 03:20 | $0.13 | $0.53 | $0.033 | openrouter |
A row is added whenever a sync detects a price change. Vendor pricing source
Pricing (USD per 1M tokens)
- Input
- $0.13
- Output
- $0.53
- Cached input
- $0.033
- Blended (3:1)
- $0.23
Details
- Company
- Tencent
- Version
- 3
- Family
- Hy
- Released
- 2026-07-06
- Parameters
- 298.8B
- Max output
- 128K
- Open weights
- Yes
- Inputs
- text
- Last synced
- 46 min ago