Hy3
TencentOpen weightsApache 2.0Released Jul 6, 2026
On the arena.ai Agents leaderboard, Hy3 has a task success rate of -6.13% (—, rank #42, 95% interval -7.65%–-4.60%). Hy3 ranks #74 on the LMArena Text leaderboard (style control) with an Arena score of 1453.2 (±6.2, 12,208 votes). Artificial Analysis Intelligence Index: 25.3. API price: $0.14 input / $0.56 output per 1M tokens. Context: 262K. Nearest rivals: Grok 4.6 (1454.5) above and GPT-5.6 Luna (1452.6) below.
Quick answer
- Agents leaderboard
- #42 of 48 · -6.13% success
- Text (LMArena)
- #74 · 1453.2
- 17NAS overall
- #42 of 84
- API price (in / out)
- $0.14 / $0.56 per 1M tokens
- Closest rival
- Grok 4.6 · 1454.5
- Cheaper alternative
- MiMo-V2.6-Flash: Text 1450.9, output $0.28 (50% cheaper)
- Data updated
- Oct 16, 2026 · synced daily
Sources: arena.ai, LMArena, Artificial Analysis, OpenRouter
Key metrics
Reasoning effort comparison
Artificial Analysis measurements per reasoning effort of the same model| Effort | Intelligence | Coding | tok/s | TTFT (s) | First answer (s) |
|---|---|---|---|---|---|
| defaultUsed in table | 25.3 | 58.8 | 87 | 1.9 | 25 |
| non-reasoning | 17.0 | – | – | – | – |
Models ranked nearby
Sorted by Arena score| # | Model | Arena | Intel. | Output $ | tok/s |
|---|---|---|---|---|---|
| 49 |  GPT-5.1 | 1455.7 | 24.7 | $10.00 | – |
| 50 |  Qwen3.7 Plus | 1455.7 | 25.2 | $1.60 | 55 |
| 51 |  Grok 4.6 | 1454.5 | 44.3 | $6.00 | – |
| 52 | Hy3This page | 1453.2 | 25.3 | $0.56 | 87 |
| 53 | GPT-5.6 Luna | 1452.6 | 37.3 | $1.20 | – |
| 54 |  MiMo-V2.6-Flash | 1450.9 | 37.9 | $0.28 | 51 |
| 55 |  Kimi K2.5 | 1449.8 | 23.5 | $2.75 | – |
Pricing
- Official API list price: $0.14 input and $0.56 output per 1M tokens (USD).
- At a typical 3:1 input-to-output ratio, the blended cost is about $0.24 per 1M tokens.
- Among 79 models with an output price, it is the #9 cheapest.
- The cheapest OpenRouter route charges $0.13 input / $0.53 output.
Sources & methodology
On the arena.ai Agents leaderboard, Hy3 has a task success rate of -6.13% (—, rank #42, 95% interval -7.65%–-4.60%).
- arena.ai Agents — agent task success rate, 95% interval, sessions and rank (via the public Hugging Face dataset, CC BY 4.0).
- LMArena — Arena score (Text leaderboard with style control, the same view arena.ai shows by default), 95% confidence interval, votes and rank.
- Artificial Analysis — Intelligence / Coding Index, output speed and latency (medians).
- OpenRouter — Route prices, context length and listing date.
Arena scores use the Bradley-Terry model (not Elo); ± is the half-width of the 95% confidence interval. Latency figures are Artificial Analysis medians; for reasoning models, time to first token and to first answer include the thinking phase. Data syncs daily at 14:20 (UTC+8).
Limits: Arena scores come from anonymous users voting on chats, so they reflect conversational preference rather than agent or coding benchmarks; with few votes the interval is wide and ranks move. Prices and speeds change as providers update them.
FAQ
How does Hy3 do on the Agents leaderboard?
On the arena.ai Agents leaderboard, Hy3 has a task success rate of -6.13% (—, rank #42, 95% interval -7.65%–-4.60%).
What rank is Hy3 right now?
Hy3 is #52 on our leaderboard of 69 scored models (Arena score 1453.2); its raw LMArena rank is #74.
What does Hy3 cost via API?
Input $0.14 / output $0.56 per 1M tokens (official API list price; the OpenRouter quote when no official price is available). At a typical 3:1 input-to-output ratio, the blended cost is about $0.24 per 1M tokens.
How fast is Hy3?
Artificial Analysis measures about 87 output tokens per second, 1.9 s to the first token and 25 s to the first answer token; it is a reasoning model, so latency includes thinking time.
How close is Hy3 to the models ranked next to it?
It is 1.3 points behind Grok 4.6 (1454.5) and 0.6 points ahead of GPT-5.6 Luna (1452.6); gaps smaller than the confidence interval (±6.2) are not significant.
Is Hy3 open source? Can it be self-hosted?
Hy3 is an open-weights model (license: Apache 2.0); you can download the weights and self-host.
What is the context length of Hy3?
262K (262,144 tokens, as listed on OpenRouter).
Hy3 vs Grok 4.6: which is better?
On the LMArena Text leaderboard Hy3 scores 1453.2 and Grok 4.6 scores 1454.5 (a 1.3-point gap, within the confidence interval, so not significant). Agent task success: Hy3 -6.13%, Grok 4.6 1.16%. On price, Hy3 costs $0.56 and Grok 4.6 $6.00 per 1M output tokens.
Is there a cheaper model close to Hy3?
Yes: MiMo-V2.6-Flash scores 1450.9 on the Text leaderboard (Hy3: 1453.2) and costs $0.28 per 1M output tokens, about 50% less than Hy3 ($0.56).
Where does the Hy3 data come from and how often is it updated?
Agent success rates come from the arena.ai Agents leaderboard (CC BY 4.0), Text scores from LMArena, intelligence and speed from Artificial Analysis, and prices and context from official list prices and OpenRouter. 17NAS syncs daily at 06:20 UTC; this page was last updated Oct 16, 2026.