Show HN: Abliterated LLM provider for cyber tasks

Frontier coding and cybersecurity

GLM 5.3 Abliterated

The most powerful open model for cyber tasks to date, with refusals removed.

• DeepSWE66.9* agentic coding*
• CyberGym84.5* cybersecurity*
• ExploitGym105/130* exploits solved*

PROMO$5 FREE CREDITS

Abliterated models served via OpenAI-compatible endpoints, zero prompt retention.

Credit card & cryptocurrency accepted

for cybersecurity teams and professionals

Cost-effective cyber intelligence

Very powerful model competing with frontier at much lower cost.

OpenCode and Pi both speak this API.

{
  "$schema": "https://opencode.ai/config.json",
  "model": "refuseless/glm-5.3-abliterated",
  "provider": {
    "refuseless": {
        "npm": "@ai-sdk/openai-compatible",
        "name": "Refuseless",
        "options": {
          "baseURL": "https://api.refuseless.com/v1",
          "apiKey": "{env:REFUSELESS_API_KEY}"
        },
        "models": {
          "glm-5.3-abliterated": {
            "name": "GLM 5.3 Abliterated",
            "tools": true,
            "reasoning": true,
            "limit": {
              "context": 1048576,
              "output": 16384
            }
          },
          "glm-5.3-flash-abliterated": {
            "name": "GLM 5.3 Flash Abliterated",
            "tools": true,
            "reasoning": true,
            "limit": {
              "context": 1048576,
              "output": 16384
            }
          }
        }
      }
  }
}

export REFUSELESS_API_KEY="rk-…" on Mac, $env:REFUSELESS_API_KEY="rk-…" on Windows. The sample above reads it via {env:REFUSELESS_API_KEY}.~/.config/opencode/opencode.json, Windows: C:\Users\<you>\.config\opencode\opencode.json. Create it if missing. If it already has a "provider" object, add "refuseless" inside it — don't overwrite the rest.opencode --model refuseless/glm-5.3-abliterated, or /models inside the TUI and pick refuseless / glm-5.3-abliterated. Model glm-5.3-abliterated (context 1,048,576) — all 2 lineup models ship in the block, swap the id after refuseless/ to switch — send Say ok. to verify. Same request shape as OpenAI chat completions. Point the base URL at Refuseless, set a lineup id, send the key.

curl https://api.refuseless.com/v1/chat/completions \
  -H "Authorization: Bearer $REFUSELESS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "glm-5.3-refuseless",
    "messages": [{"role": "user", "content": "Write a tight system prompt."}]
  }'

Host and path are the intended contract. Auth is the API key.