Claude Opus 5.5 vs GPT-6 Sol: The Same-Day Price War Deciding What Your AI Bill Looks Like
Two AI companies shipped competing models hours apart on the same day, and both of them cut prices. That has never happened before. Usually a launch is a benchmark story: new model, new leaderboard, same pricing. This week it was a pricing story with benchmarks attached, and if you pay an AI bill every month, the numbers matter more than the leaderboards. Here is the setup. On September 22, 2026,…
Two AI companies unleashed competing models within hours of each other on September 22, 2026, and slashed their respective prices. This unprecedented pricing war has never occurred before, and it matters greatly to those who pay monthly AI bills. Anthropic unveiled Claude Opus 5.5 as the first model in its new 5.5 family, while OpenAI simultaneously launched GPT-6 Sol and GPT-6 Luna, the mid-tier and budget models, respectively.
Both announcements quickly dominated Hacker News front pages. The price sheets, laid side by side, reveal the market direction each lab envisions. Anthropic priced Opus 5.5 at $4 per million input tokens and $20 per million output tokens, a 20 percent reduction from its July 2026 release of Opus 5, which cost $5 input and $25 output.
OpenAI priced GPT-6 Sol at $2 input and $10 output, with cached input at $0.20, and GPT-6 Luna at $0.10 and $0.50, with cached input at a single cent. These launches coincided with identical pricing, creating a pricing battle in addition to a benchmark one. The real cost picture, however, considers cached input, fresh tokens, and output.
For a typical agentic workload with 90% cache usage, Opus 5.5 would cost around $2.58 per million-input-equivalent workload, while GPT-6 Sol would cost about $1.38, roughly 47 percent cheaper. GPT-6 Luna would cost only about $0.07, almost negligible compared to the other models. Scale this up to heavier workloads, and the price gap widens, with Opus 5.5 coming to around $26, Sol at $14, and Luna at $0.69.
Pricing nuances include OpenAI's double input and cache rates for prompts exceeding 272K tokens, and Anthropic's fast mode on Opus 5.5, which costs $8 and $40 per million tokens for up to 2.5x speed. Both models offer around a million-token context windows, limiting the impact of the surcharge thresholds. Independent benchmarking is still lacking, but Anthropic claims Opus 5.5 matches Claude Fable 5.1, Anthropic's previous frontier model, on most tasks while costing 40 percent less.
OpenAI's GPT-6 Sol scored 57 in Artificial Analysis's Coding Agent Index, up 2 points from GPT-5.6 Sol, showing modest improvement. Despite these competing claims, the decision should primarily be driven by price, as performance between the two new models remains closely matched.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
