GetRun / Chat & Text
AI Chat Models Through One API
GetRun serves chat and text models from several providers behind one OpenAI-compatible endpoint: POST /v1/chat/completions. An OpenAI SDK you already use only needs its base_url pointed at GetRun, with no rewriting. Billing is per token in rupiah, with separate rates for input, cached input, and output. Because every model sits behind one endpoint, moving between providers — Claude, GPT, or Grok — means changing a single "model" value in the request body rather than rewriting the integration. That is also what makes comparing cost across models possible on exactly the same workload.
There are currently 15 chat & text models from 4 providers, with sample runs starting at Rp2,784 each.
Why through GetRun
- An OpenAI-compatible endpoint: an existing SDK only needs a new base_url and API key.
- A rupiah price per million tokens, with different rates for input, cached input, and output.
- Streaming is supported; a streamed response closes with a single event carrying the task_id and the final price.
- The final cost is computed from tokens actually used, not from the estimate.
- Also available as an MCP connector, so MCP clients can spend a GetRun balance without an API key.
- Models from several providers at once — Claude (Anthropic), GPT (OpenAI), and Grok (xAI) — behind one endpoint and one API key.
- Switching models means changing one "model" value in the request body; the rest of the integration is untouched.
- The same prepaid balance covers chat as well as image, video, and audio models, so there is no separate bill per provider.
How to use it
- Get an API key — Create an API key in the dashboard, then send it as the Authorization: Bearer header.
- Point your SDK at GetRun — Set base_url to the GetRun endpoint and keep the rest of your OpenAI SDK code as it is.
- Watch the cost — Each response carries the task_id and the final price, computed from the tokens actually used.
Available chat & text models
- GPT-5.6 Sol by OpenAI — Rp24,750 / 1M token input
- GPT-5.6 Terra by OpenAI — Rp12,375 / 1M token input
- GPT-5.6 Luna by OpenAI — Rp4,950 / 1M token input
- GPT-5.5 by OpenAI — Rp24,750 / 1M token input
- GPT-5.4 by OpenAI — Rp12,375 / 1M token input
- GPT-5.4 Mini by OpenAI — Rp3,713 / 1M token input
- GPT-5.3 Codex Spark by OpenAI — Rp8,663 / 1M token input
- Claude Opus 5 by Anthropic — Rp24,750 / 1M token input
- Claude Opus 4.8 by Anthropic — Rp24,750 / 1M token input
- Claude Sonnet 5 by Anthropic — Rp10,519 / 1M token input
- Claude Fable 5 by Anthropic — Rp49,500 / 1M token input
- Grok 4.7 by xAI — Rp9,900 / 1M token input
- Grok 4.6 by xAI — Rp9,900 / 1M token input
- Grok 4.5 by xAI — Rp9,900 / 1M token input
- Gemini 3.7 Flash by Google — Rp2,784 / 1M token input
Frequently asked questions
- Is the GetRun chat API compatible with the OpenAI SDK?
- Yes. POST /v1/chat/completions follows the same shape, so an existing SDK only needs its base_url and API key changed.
- How is chat billing calculated?
- Per token in rupiah, with separate rates for input, cached input, and output. The final figure comes from tokens actually used, not from an estimate.
- Is streaming supported?
- Yes. A streamed response closes with one event carrying the task_id and the final price.
- How is this different from the MCP connector?
- The API is for your own code with an API key. The MCP connector lets an MCP client spend your GetRun balance through OAuth, without copying a key anywhere.
- Which providers are available for chat?
- Claude from Anthropic, GPT from OpenAI, and Grok from xAI, all behind the same endpoint and the same API key.
- How do I switch from one chat model to another?
- Change the "model" value in the request body. Nothing else in the integration changes.
- Do I need a monthly subscription for chat models?
- No. Chat draws on the same prepaid rupiah balance as every other model, with nothing renewing automatically.
Other categories
Standards we follow
Full catalog · Price list · Compare models · API documentation · About GetRun