Kimi K3 vs Qwen3-VL-2B-Instruct

Kimi K3

Kimi K3 is a 2.8T multimodal agent model with a 1M-token context, parallel tool calls and per-request effort, for long coding and research sessions. Its 1,561 GB of files need a multi-node GPU cluster, so most agents will call a hosted copy. The main risks are Moonshot's Python running at load and prompt time, and no answers in its repository.

Qwen3-VL-2B-Instruct

The smallest Qwen3-VL cousin, for reading screenshots, documents, charts and video on a laptop or edge box, running in about 2 GB from Alibaba's Q4_K_M GGUF in every major runtime under plain Apache 2.0. Qwen hasn't answered repository questions or published safety work, so keep a human or a check in the loop before it drives a GUI.

Kimi K3

1. Start it with SGLang on a node of 8 B300s, sglang serve --trust-remote-code --model-path moonshotai/Kimi-K3 --tp-size 8 --context-length 65536 --mem-fraction-static 0.82 --reasoning-parser kimi_k3 --tool-call-parser kimi_k3 --port 30000
2. Use the kimi_k3 tool-call and reasoning parsers in vLLM or SGLang, and validate tool-call arguments against the schema, since vLLM's recipe says K3 sometimes emits a format its parser doesn't expect
3. Pass the whole assistant message back every turn, reasoning_content and tool_calls included, or output becomes unstable
4. Thinking is always on; set reasoning_effort to low or high when max is too slow
5. Sampling per the card is temperature 1.0, with top-p 0.95 for single-step tasks and 1.0 for agentic ones
6. Don't hand an ongoing session from another model to K3, and state boundaries in the system prompt, since Moonshot warns it can act on its own when intent is unclear
7. Without the hardware, call it through Moonshot's API or an inference provider listed on its Hugging Face page

Qwen3-VL-2B-Instruct

1. Start it with llama-server -hf Qwen/Qwen3-VL-2B-Instruct-GGUF:Q4_K_M --jinja -c 32768, or ollama run qwen3-vl:2b-instruct
2. In Ollama the plain qwen3-vl:2b tag is the Thinking edition, so name 2b-instruct
3. With the GGUF, load the mmproj vision file too, or images won't be read
4. Parse tool calls as JSON inside <tool_call> tags, or use SGLang's --tool-call-parser qwen
5. Use the card's sampling, temperature=0.7, top_p=0.8, top_k=20, presence_penalty=1.5 for vision and temperature=1.0, top_p=1.0, top_k=40, presence_penalty=2.0 for text
6. Validate JSON output yourself, since structured output isn't documented
7. vLLM needs 0.11.0 or later and Transformers 4.57.0 or later