Hy4 Xploded in OpenCode

Hy4 Xploded in OpenCode

Run hy4-xploded in OpenCode through MagmaRouter: one key, a hard spend cap, and a model that answers instead of refusing.

Model idhy4-xploded
Context window1,048,576 tokens (1M)
Tool callingSupported
Price per 1M tokens$3.20 input, $0.60 on a cache hit, $6.40 output
Web searchAdd :online to the id

Set it up

  1. Add a magmarouter provider to opencode.json with this model and its limits.
  2. Export MAGMAROUTER_API_KEY in the shell OpenCode starts from.
  3. Run opencode and pick magmarouter/hy4-xploded with /models.
{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "magmarouter": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "MagmaRouter",
      "options": {
        "baseURL": "https://api.magmarouter.com/v1",
        "apiKey": "{env:MAGMAROUTER_API_KEY}"
      },
      "models": {
        "hy4-xploded": {
          "name": "Hy4 Xploded",
          "limit": { "context": 1048576, "output": 32768 }
        }
      }
    }
  },
  "model": "magmarouter/hy4-xploded",
  "small_model": "magmarouter/hy4-xploded"
}
OpenCode with Hy4 Xploded OpenCode plans its context from limit.context. Leave it at 1,048,576 or it compacts the session long before the model would have needed it.

Hy4 Xploded has its own page with every limit and price.

What a session costs

For a 25-turn agent task (about 750K input tokens re-read across turns, 85% of them cache hits, 25K output), Hy4 Xploded comes to about $0.90 at today's prices; the cache price does most of the work, since re-read context is billed at $0.60 rather than $3.20 per million.

FAQ

Can Hy4 Xploded edit files and run commands in OpenCode?

Yes. OpenCode acts through tool calls and this model supports tool calling, so it can read, edit and run, not just chat.

What does a session cost?

About $0.90 for a 25-turn agent task (about 750K input tokens re-read across turns, 85% of them cache hits, 25K output), at $3.20 in, $6.40 out and $0.60 per million on cache hits. Put a spend cap on the key so a long run cannot overrun it.

Is it uncensored?

Tencent's Hy4, uncensored: a million tokens of context, tool calling, reasoning and vision.

The other model in OpenCode