Benchmark scores are exciting but more importantly we are seeing incredible results so far using this model in Devin! Try it out:
Introducing SWE-1.7, the most capable model we’ve trained yet.
It scores within a few points of the strongest frontier models at a fraction of the cost, and is now available at 1000 tok/s.
RL is not hitting its limit: after refining our recipe, we keep seeing gains as we scale
もっと見る
