{"version": 2, "width": 120, "height": 32, "timestamp": 1777905050, "idle_time_limit": 2.0, "env": {"SHELL": "/bin/bash", "TERM": "xterm-256color"}, "title": "Kiln: drop-in OpenAI-compatible chat"}
[0.00611, "o", "$ "]
[0.407618, "o", "."]
[0.427452, "o", "/"]
[0.446602, "o", "t"]
[0.465893, "o", "a"]
[0.485202, "o", "r"]
[0.504701, "o", "g"]
[0.524377, "o", "e"]
[0.543974, "o", "t"]
[0.562778, "o", "/"]
[0.581552, "o", "r"]
[0.601768, "o", "e"]
[0.620912, "o", "l"]
[0.640003, "o", "e"]
[0.658846, "o", "a"]
[0.677752, "o", "s"]
[0.697853, "o", "e"]
[0.717317, "o", "/"]
[0.737105, "o", "k"]
[0.756025, "o", "i"]
[0.775023, "o", "l"]
[0.794055, "o", "n"]
[0.813471, "o", " "]
[0.833711, "o", "s"]
[0.853331, "o", "e"]
[0.873392, "o", "r"]
[0.893471, "o", "v"]
[0.912849, "o", "e"]
[0.933106, "o", " "]
[0.952289, "o", "-"]
[0.972376, "o", "-"]
[0.991366, "o", "c"]
[1.010564, "o", "o"]
[1.030778, "o", "n"]
[1.050898, "o", "f"]
[1.069942, "o", "i"]
[1.088984, "o", "g"]
[1.108335, "o", " "]
[1.12859, "o", "k"]
[1.148716, "o", "i"]
[1.167752, "o", "l"]
[1.186946, "o", "n"]
[1.206399, "o", "."]
[1.226087, "o", "e"]
[1.245439, "o", "x"]
[1.264885, "o", "a"]
[1.285148, "o", "m"]
[1.30537, "o", "p"]
[1.324613, "o", "l"]
[1.34447, "o", "e"]
[1.363618, "o", "."]
[1.382676, "o", "t"]
[1.401917, "o", "o"]
[1.42175, "o", "m"]
[1.441893, "o", "l"]
[1.460972, "o", " "]
[1.480285, "o", "&"]
[1.499145, "o", "\r\n"]
[5.042929, "o", "$ "]
[5.444066, "o", "c"]
[5.462787, "o", "a"]
[5.481507, "o", "t"]
[5.500381, "o", " "]
[5.519675, "o", "d"]
[5.538626, "o", "o"]
[5.557527, "o", "c"]
[5.576401, "o", "s"]
[5.595185, "o", "/"]
[5.615644, "o", "s"]
[5.634813, "o", "i"]
[5.653592, "o", "t"]
[5.67245, "o", "e"]
[5.691232, "o", "/"]
[5.711459, "o", "d"]
[5.731531, "o", "e"]
[5.751467, "o", "m"]
[5.770329, "o", "o"]
[5.790122, "o", "/"]
[5.810243, "o", "d"]
[5.829574, "o", "e"]
[5.848468, "o", "m"]
[5.867216, "o", "o"]
[5.88604, "o", "-"]
[5.905057, "o", "o"]
[5.923944, "o", "p"]
[5.942929, "o", "e"]
[5.961739, "o", "n"]
[5.980673, "o", "a"]
[5.999388, "o", "i"]
[6.018571, "o", "."]
[6.037913, "o", "p"]
[6.056753, "o", "y"]
[6.075548, "o", "\r\n"]
[6.07637, "o", "\"\"\"OpenAI Python client drop-in demo for Kiln.\r\n\r\nRun via demo-openai.sh, which records the asciicast. The script itself just\r\npoints the official `openai` client at http://localhost:8420/v1 and streams a\r\nchat completion — no Kiln-specific code at all.\r\n\"\"\"\r\nfrom openai import OpenAI\r\n\r\nclient = OpenAI(base_url=\"http://localhost:8420/v1\", api_key=\"kiln\")\r\n\r\nstream = client.chat.completions.create(\r\n    model=\"Qwen3.5-4B\",\r\n    messages=[\r\n        {\"role\": \"user\", \"content\": \"In two short sentences, why does single-GPU LLM inference matter?\"},\r\n    ],\r\n    max_tokens=96,\r\n    temperature=0.3,\r\n    seed=5,\r\n    stream=True,\r\n)\r\n\r\nfor chunk in stream:\r\n    delta = chunk.choices[0].delta.content\r\n    if delta:\r\n        print(delta, end=\"\", flush=True)\r\nprint()\r\n"]
[8.07758, "o", "$ "]
[8.478745, "o", "p"]
[8.497983, "o", "y"]
[8.518344, "o", "t"]
[8.537461, "o", "h"]
[8.556549, "o", "o"]
[8.575483, "o", "n"]
[8.594587, "o", "3"]
[8.614717, "o", " "]
[8.634082, "o", "d"]
[8.652954, "o", "o"]
[8.671742, "o", "c"]
[8.690597, "o", "s"]
[8.71019, "o", "/"]
[8.729243, "o", "s"]
[8.74938, "o", "i"]
[8.768332, "o", "t"]
[8.787507, "o", "e"]
[8.807658, "o", "/"]
[8.826684, "o", "d"]
[8.845696, "o", "e"]
[8.864691, "o", "m"]
[8.883476, "o", "o"]
[8.90225, "o", "/"]
[8.921711, "o", "d"]
[8.940768, "o", "e"]
[8.959734, "o", "m"]
[8.978481, "o", "o"]
[8.99725, "o", "-"]
[9.017578, "o", "o"]
[9.036857, "o", "p"]
[9.055744, "o", "e"]
[9.074503, "o", "n"]
[9.093336, "o", "a"]
[9.113752, "o", "i"]
[9.132982, "o", "."]
[9.151874, "o", "p"]
[9.170623, "o", "y"]
[9.189433, "o", "\r\n"]
[10.193058, "o", "Single"]
[10.202129, "o", "-G"]
[10.230362, "o", "PU"]
[10.256959, "o", " inference"]
[10.282792, "o", " matters"]
[10.309996, "o", " because"]
[10.335739, "o", " it"]
[10.362139, "o", " drastically"]
[10.38936, "o", " reduces"]
[10.417161, "o", " latency"]
[10.443689, "o", " and"]
[10.468976, "o", " costs"]
[10.496184, "o", " by"]
[10.522989, "o", " eliminating"]
[10.549601, "o", " the"]
[10.574666, "o", " need"]
[10.601501, "o", " for"]
[10.627299, "o", " expensive"]
[10.6538, "o", " multi"]
[10.679862, "o", "-G"]
[10.706488, "o", "PU"]
[10.732241, "o", " clusters"]
[10.75892, "o", " or"]
[10.784276, "o", " complex"]
[10.810377, "o", " cloud"]
[10.836125, "o", " orchest"]
[10.862788, "o", "ration"]
[10.89015, "o", "."]
[10.919943, "o", " This"]
[10.945364, "o", " approach"]
[10.971726, "o", " enables"]
[10.998583, "o", " faster"]
[11.024929, "o", ","]
[11.051543, "o", " more"]
[11.077507, "o", " responsive"]
[11.103428, "o", " interactions"]
[11.130706, "o", " for"]
[11.156571, "o", " end"]
[11.183639, "o", "-users"]
[11.210256, "o", " and"]
[11.236207, "o", " allows"]
[11.262458, "o", " developers"]
[11.289561, "o", " to"]
[11.317365, "o", " deploy"]
[11.342714, "o", " models"]
[11.368865, "o", " locally"]
[11.396335, "o", " without"]
[11.422923, "o", " significant"]
[11.449208, "o", " infrastructure"]
[11.475646, "o", " overhead"]
[11.502992, "o", "."]
[11.5313, "o", "\r\n"]