qwen3.8-27b
Qwen3.8-27B is an open-weights (Apache 2.0) dense model from the Qwen team, served by CedarLLM as a 4-bit build with a 262,144-token context. It handles chat, coding, long-document analysis, image understanding (photos, screenshots, scanned pages), tool/function calling and JSON output, with optional step-by-step thinking for math and logic. Call it through the OpenAI-compatible endpoint https://api.cedarllm.com/v1/chat/completions with model "qwen3.8-27b" and a Bearer API key. Use cases: assistants, agents, RAG over long documents, data extraction, OCR-style reading and coding help.