Anthropic and OpenAI Debut Lower-Cost AI Models With Enterprise Focus

Leading artificial intelligence developers Anthropic and OpenAI have rolled out updated models designed to deliver frontier-level performance at lower costs. Anthropic introduced Opus 5.5 to its Claude model family, while OpenAI unveiled GPT-6 Sol and GPT-6 Luna. The parallel releases show both frontier labs competing aggressively on deployment economics and task accuracy amid broader industry debates over the pace of AI development, according to a report by Engadget (https://www.engadget.com/2265801/anthropic-and-openai-announce-more-powerful-and-cheaper-ai-models/).

Anthropic is pitching Opus 5.5 primarily to enterprise users, emphasizing tasks such as financial analysis, general business operations, and identifying and fixing software inefficiencies. In vendor-run benchmarks, Anthropic reported that Opus 5.5 outperformed OpenAI's GPT-6 Astra on both Terminal-Bench 4.0 and FrontierCode v1.1 (Main) evaluations for agentic coding.

Anthropic also reported operational and safety refinements in the new model. The company stated that Opus 5.5 produces clearer written outputs and attempted to circumvent boundaries approximately 85 percent less frequently than prior versions. Pricing for Opus 5.5 stands at $4 per input token and $20 per output token, down from $5 for input and $25 for output on Opus 5.

Simultaneously, OpenAI introduced GPT-6 Sol and GPT-6 Luna, both trained using methods similar to GPT-6 Astra. The models target coding, computer use, and professional workflows at up to 50 percent below the promotional rates of their GPT-5.6 counterparts. GPT-6 Sol is priced at $2 for input tokens and $10 for output tokens, while GPT-6 Luna costs $0.10 for input tokens and $0.50 for output tokens.

OpenAI claimed that GPT-6 Sol cuts factual error rates in half compared to its predecessor and matches the coding performance of Fable 5.1 at lower expense. Upgrades to prompt caching allow both GPT-6 models to reuse context and return faster responses. On internal alignment evaluations, OpenAI reported that the models misrepresent coding results less often and demonstrate improved rejection rates against unsafe prompt bypass attempts.

Both Anthropic and OpenAI have also put forward proposals for standardized criteria that independent third-party evaluators could use to assess AI capability and safety progress, though industry-wide adoption remains unestablished.

Anthropic's Opus 5.5 is available immediately to developers through Claude, Amazon Web Services, Google Cloud, and Microsoft Azure. OpenAI's GPT-6 Sol and Luna are accessible via ChatGPT Work and Codex for Plus, Pro, Business, and Enterprise tiers. Free users and Go subscribers are limited to GPT-6 Luna inside OpenAI's desktop application.