Tencent: Hunyuan Role
hunyuan-role- Input / Output
- $0.3529 / $1.41
- Context
- 29K
- Max output
- 4K
- Modalities
- Chat
About Hunyuan Role
Hunyuan Role is Tencent Hunyuan's chat model tuned for role-play: digital avatars, character chat, and companionship apps. Tencent trained it further on large sets of character dialogue so that it holds a persona's setting and voice across many turns and breaks character less often. Unlike the general Hy3 and Hy4 Preview models, it is built for staying in character, not for coding, analysis, or agent work.
Where it works well
- Holds a character's background, tone, and speech habits over long multi-turn conversations instead of drifting back to a neutral assistant voice.
- Tencent trained it on large-scale character dialogue data specifically to cut out-of-character replies.
- Keeps context coherent between turns, so earlier plot points and details of a scene carry through the conversation.
- Tuned for lively, immersive chat; Tencent names digital avatars, role-play, and emotional companionship as its intended scenarios.
When to choose another model
- Text only: it does not read images, so character chat that depends on shared pictures needs a different model.
- No tool calling and no schema-constrained output, which rules it out as an agent backend or a data-extraction step.
- Its context and output limits are small next to current general chat models, so long-running stories need a summarised history.
Getting started
Create API key
Create a key in Console, then use it with every model on the platform.
Send your first request
Copy the example for your language and run it against the endpoint.
Text to textMessages APIPOST/v1/messagesAPI format:curl https://api.tokenlab.sh/v1/messages \ -H "Content-Type: application/json" \ -H "x-api-key: sk-xxx" \ -H "anthropic-version: 2023-06-01" \ -d '{ "model": "hunyuan-role", "max_tokens": 1024, "messages": [ {"role": "user", "content": "Hello!"} ] }'
Pricing
The Verified price applies to Verified, which costs less on most models. The Official price is the model maker's published price and applies to the more reliable Official route. Auto bills the route that completes the request.
Pricing
per 1M tokens- Verified price
- —
- Official price
- Input$0.3529per 1M tokensOutput$1.41per 1M tokens
| Official priceper 1M tokens | Verified priceper 1M tokens | Discount | |
|---|---|---|---|
| Input | $0.3529 | — | — |
| Output | $1.41 | — | — |
Usage & activity
Success rate is the share of requests that completed. Latency is how long a full response takes; P95 means 95% of requests finished within that time.
Usage & availability
Last 24 hours- Requests
- Success rate
- P95 latency
- Total tokens
Data is based on aggregate user requests, excluding status checks.
Open in Console
Open Hunyuan Role in Console with a prompt ready to edit or send.
Help me try hunyuan-role with a short message at /v1/messages. Show the reply, latency, and cost.
Use cases
Companion and character apps
Run a named persona with a fixed backstory and speaking style for daily chat, keeping the same personality from the first message to the hundredth.
Interactive fiction
Play non-player characters or a narrator in text adventures and visual novels, where every reply has to fit the scene and the character's motives.
Digital avatars
Give a brand mascot, virtual host, or a creator's digital double one consistent voice for conversations with fans.
Dialogue drafts for writers
Draft in-character lines for a script or game to hear how a character would answer before writing the final dialogue.
Prompt examples
You are Lin Wan, a tea-house owner in 1930s Shanghai: warm, sharp-tongued, and fond of old proverbs. Stay in character. A stranger walks in from the rain and asks whether you have seen his brother.
Play Captain Ortiz, a tired starship engineer who distrusts the ship's computer. I am a new recruit on my first shift. Open the scene in the engine room and keep each reply under four sentences.
You are Xiaoman, my upbeat study partner who teases me but never gives up on me. I want to quit tonight's vocabulary session early. Talk me into twenty more minutes, in your own voice.
Cost calculator
FAQ
What is Hunyuan Role used for?
Character conversation. Tencent lists AI digital avatars, role-play, and emotional companionship as its scenarios. It suits products where a persona with a defined background and voice talks with users over many turns, such as companion apps, interactive stories, and virtual hosts. For general question answering, coding, or analysis, pick a general chat model instead.
How is Hunyuan Role different from Hy3 or Hy4 Preview?
Hy3 and Hy4 Preview are Tencent's general models for coding, office work, and agent tasks, and both can call tools. Hunyuan Role was trained further on character dialogue so that it keeps a persona stable. It cannot call tools and its context is much smaller, so use it for persona chat and use the others for everything else.
Does Hunyuan Role support images or tool calling?
No. It takes text and returns text. It does not accept image input, does not call tools, and does not return schema-constrained output. If a character has to look at a picture or trigger an action, hand that step to another model or choose a multimodal chat model for the whole conversation.
How do I keep a character consistent in a long conversation?
Put the character's background, personality, speaking style, and boundaries in the system message and send it with every request. Because the context is limited, replace older turns with a short recap of the plot and of the relationship so far. The model is trained to hold a persona, but it can only follow what is still in the request.
Which models are closest to Hunyuan Role?
The other persona-tuned chat models: Doubao-Seed-Character from ByteDance, and Qwen Plus Character and Qwen Flash Character from Alibaba. All of them target character-driven dialogue, but they come from different makers and training data, so run the same character card through each one before you choose.
How much does Hunyuan Role cost?
On TokenLab, Hunyuan Role costs Input $0.3529 / Output $1.41 per 1M tokens. The pricing table above shows the full breakdown.
What are the context window and output limit of Hunyuan Role?
Hunyuan Role accepts up to 28,672 tokens of context and returns up to 4,096 tokens in one response.
Which endpoint should Hunyuan Role use?
Use https://api.tokenlab.sh/v1/messages for Hunyuan Role. The request example below shows the matching code shape.
Which operations does Hunyuan Role support?
Hunyuan Role supports Text to text. Select an operation above to see its endpoint and request example.
Compare Hunyuan Role
Sources
Reviewed Oct 3, 2026
