Higgs Realtime

Higgs Realtime passed 46 of 82 scenarios on all three runs, 56.1%, 10th of 14 realtime models. Across all 246 calls, 64.6% passed every check.

Higgs Realtime starts replying 1.94s after the caller finishes at the median, 2.02s at P90 and 2.08s at P95. That is 8th fastest of 14 at the median, measured on the call audio.

Higgs Realtime falls furthest behind the field on Medicare compliance and routing: 0% of calls passed, against a median of 48%. When it saved data wrong, the most common mistake was step skipped, 39% of them.

Higgs Realtime saved everything exactly right on 66.2% of the calls that save data, 12th of 14. On Booking calls, 91.8%. On Medicare calls, 0.0%.

$0.0061 a minute of call at Boson AI's published rates, 1st cheapest of 10 priced models. Boson's published token rates applied to each call's reported usage; caller transcription ($0.0025 a minute) is not in the usage the service reports, so it is not included.

Cekura ran Higgs Realtime as the whole agent in the same open-source Pipecat pipeline as every other model, with the same prompt, tools and connection. A simulated caller worked through 82 scenarios on live calls, three times each: 246 calls for Higgs Realtime. A call passes only when the agent said the right things and saved the right data.