gpt-oss-120b vs Qwen3-4B-Instruct-2507

gpt-oss-120b

A reasoning and agent cousin that runs in most runtimes under Apache 2.0 with published safety work. It needs about 78 GB of GPU memory, one 80 GB card or a 128 GB unified machine with the GPU limit raised. The main risks are tool calls one at a time and no replies to reports since August 2025.

Qwen3-4B-Instruct-2507

A small, fast cousin for instruction following, short tool-calling jobs and long documents on modest hardware, with no thinking step. It runs in about 3 GB at 4 bits on every common runtime, and two providers host it. The main risk is neglect, with no file changes in over a year, no answers from Alibaba and no safety results.

gpt-oss-120b

1. Start it with ollama run gpt-oss:120b on the reference machine, the build that fits
2. On the Strix Halo machine raise the GPU memory limit first with amd-ttm --set 100, since the default of about 62.8 GiB is below the 78 GB it needs
3. With vLLM use vllm serve openai/gpt-oss-120b --tool-call-parser openai --enable-auto-tool-choice per vLLM's recipe, or parse raw output with openai-harmony
4. Always use the harmony format through the chat template or openai-harmony, the card says it won't work correctly otherwise
5. Set Reasoning: low in the system message for short tasks, medium is the default
6. Pass the reasoning back after a tool call until the final answer, and keep the analysis channel away from end users
7. Expect one tool call per turn and loop, don't wait for parallel calls

Qwen3-4B-Instruct-2507

1. Quickest start is ollama run qwen3:4b-instruct, 2.5 GB. Plain qwen3:4b is the Thinking-2507 edition, not this one
2. On our reference machine use BF16 with vllm serve Qwen/Qwen3-4B-Instruct-2507 --max-model-len 262144 --enable-auto-tool-choice --tool-call-parser hermes
3. Use the hermes parser in vLLM or qwen25 in SGLang, the template emits JSON inside tool_call tags
4. Don't pass enable_thinking, this edition never thinks and the card says the flag isn't needed
5. Sampling per the card is temperature 0.7, top_p 0.8, top_k 20 and min_p 0, with presence_penalty up to 2 against repetition
6. Drop the context to 32,768 tokens if memory runs out, as the card suggests
7. Validate JSON yourself or use the runtime's guided decoding, since structured output isn't documented