Home › Models › Qwen2.5 72B Instruct

Qwen2.5 72B Instruct API

Open-weights economics, served on Bridge's own capacity: durable low pricing and the lowest measured latency in the catalog — the workhorse for high-volume pipelines.

Why developers pick it

• Lowest p50 latency on Bridge — measured by our 5-minute health probe, published on the models page.
• Open weights: pricing isn't tied to a closed lab's rate card.
• Ideal for classification, summarization, extraction and translation at scale.

Code example

curl https://bridge.computeindex.net/v1/chat/completions \
  -H "Authorization: Bearer sk-bridge-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen2.5-72b-instruct",
    "messages": [
      {"role": "system", "content": "Classify support tickets as: billing, bug, feature, other."},
      {"role": "user", "content": "My invoice shows a double charge for June."}
    ]
  }'

FAQ

Is this the same open model I can download myself? Yes — the open-weight Qwen2.5 72B Instruct, served with an OpenAI-compatible API and per-token metering.

When should I use it instead of DeepSeek V4 Pro? For simple, high-volume tasks where price-per-token and latency dominate; keep DeepSeek for hard reasoning and code.

Related