This matters because it gives a fast third-party read on GPT-5.5 beyond launch-day marketing. Arena says GPT-5.5 landed at #2 in Search Arena, #5 in Expert Arena, and #9 in Code Arena with a 50-point gain over GPT-5.4.
This matters because the next bottleneck in agent coding is human attention, not raw model speed. OpenAI says Symphony lifted landed pull requests by 500% on some teams after engineers hit a practical ceiling of roughly three to five concurrent Codex sessions.
One of AI’s most important commercial contracts just loosened up. Microsoft keeps Azure’s first-stop role and long-dated IP access, but OpenAI can now sell across any cloud and Microsoft will no longer pay it a revenue share.
Qualcomm $QCOM rose to $148.85, up $15.01 or about 11.2%, after analyst Ming-Chi Kuo said OpenAI is working with Qualcomm and MediaTek on smartphone processors for a 2028 device. None of the companies confirmed the report, but the move put Qualcomm back at the center of the AI hardware trade.
Why it matters: public coding benchmarks are getting less useful at the frontier, so a fresh product-side score can move developer attention fast. Cursor says GPT-5.5 is now its top model on CursorBench at 72.8% and is discounting usage by 50% through May 2.
HN did not greet GPT-5.5 with applause first. The thread went straight to pricing, context tiers, and whether the model actually behaves better once real coding work starts.
Why it matters: API availability is the moment a flagship model becomes something teams can actually wire into products. OpenAI’s developer account says GPT-5.5 brings fewer retries, and the official release page now lists API access with a 1M context window and updated pricing.
OpenAI is pushing harder into agentic work, not just chat. On the company's own evals, GPT-5.5 reaches 82.7% on Terminal-Bench 2.0, beats GPT-5.4 by 7.6 points, and uses fewer tokens in Codex.
OpenAI is pitching GPT-5.5 as more than a routine model refresh. With 82.7% on Terminal-Bench 2.0, 58.6% on SWE-Bench Pro, and a claim that it keeps GPT-5.4-level latency, the company is resetting expectations for long-running coding agents.
OpenAI is moving from generic chat to a healthcare-specific workspace, and the timing is clear: 72% of physicians now report AI use in clinical practice. The new product is free to verified U.S. physicians, NPs, PAs, and pharmacists, and OpenAI says doctors rated 99.6% of tested responses safe and accurate across 6,924 conversations.
Privacy tooling usually breaks at scale or forces raw text onto a server. OpenAI’s 1.5B open-weight Privacy Filter runs locally, handles 128,000-token inputs, and posts 97.43% F1 on a corrected PII-Masking-300k benchmark.
OpenAI is attaching cash to the hardest kind of safety failure: a single prompt that breaks all five of its bio safeguards. The new GPT-5.5 Bio Bug Bounty pays $25,000 for a universal jailbreak, limits testing to GPT-5.5 in Codex Desktop, and starts formal testing on April 28.