Skip to content
LLM Pulse

Leaderboard

τ²-Bench Telecom

5 models scored · metric: success rate. Prices are the cheapest listed across all serving providers. Pts/$ = score ÷ blended price (3×input + output) / 4 — how much score each dollar buys.

Best value · top 4 by pts per dollar
  1. 1.GPT-5.4 nano226
  2. 2.Grok 4.362.7
  3. 3.GPT-5.4 mini61.9
  4. 4.GPT-5.59.6
#ModelScoreBest in /MBest out /MPts / $Context
🥇GPT-5.5openai
98
$4.55$27.279.61.05M
🥈Grok 4.3xai
98
$1.25$2.5062.71M
🥉GPT-5.4 miniopenai
93.4
$0.68$461.9400K
4GPT-5.4 nanoopenai
92.5
$0.18$1.10226400K
5North Mini Codecohere
37
256K

Cheapest scorer: GPT-5.4 nano at $0.18 input /M.