{"object":"list","data":[{"id":"qwen3.8-27b","name":"Qwen3.8 27B (int4)","created":1787313600,"context_length":150000,"max_output_tokens":65536,"quantization":"int4","modalities":{"input":["text"],"output":["text"]},"pricing":{"prompt":"0.00000034","completion":"0.00000255","input_cache_read":"0.000000034","input":0.34,"output":2.55,"cached_input":0.034},"supported_parameters":["temperature","top_p","top_k","max_tokens","stop","seed","frequency_penalty","presence_penalty","logprobs","top_logprobs","stream","stream_options","tools","tool_choice","response_format"],"datacenter_locations":["CZ"],"compliance":{"zdr":true,"hipaa":false},"retention":"zero-retention: prompts and outputs are never persisted","is_ready":true,"hugging_face_id":"Qwen/Qwen3.8-27B","description":"Qwen3.8-27B served on vLLM at W4A16 int4, text in / text out. 150,000-token context, 65,536 max output. Streaming with usage accounting, tool calling, structured outputs, logprobs and seed. Zero data retention: prompts and completions are never written to disk, logs or analytics. Single-node capacity in the Czech Republic; we return an early 429 at saturation rather than queueing.","openrouter":{"slug":"qwen/qwen3.8-27b"},"object":"model","owned_by":"jetinfer"}]}