Anthropic iyo OpenAI wada soo saareen nidaamyo cusub oo lacagta hooseeya
Shirkadaha AI-ka ee Anthropic iyo OpenAI ayaa maanta soo saareen nidaamyo cusub oo ay ku xoojiyeen kharashka iyo kalsoonida, iyagoo diiradda saaray horumarro aad u yar laakiin muhiim u ah.
Opus 5.5: Nidaamka cusub ee Anthropic
Anthropic ayaa soo saartay Opus 5.5, nidaamkeeda ugu dambeeyay ee shaqooyinka aadka u adag sida barnaamijyada iyo xogta. Nidaamkan cusub wuxuu leeyahay kharash aad u hooseeya, iyadoo input-ka iyo output-ka ay ku baxaan $4 iyo $20 bishii.
Opus 5.5 wuxuu sameeyaa shaqooyinka 30% ka badan Opus 5, halka kharashka cache-ka uu hoos u dhacay 60%. Tani waxay muujisaa in nidaamkan cusub uu yahay mid aad u wanaagsan oo lacagta hooseeya.
GPT-6 Sol iyo Luna: Nidaamyada OpenAI
OpenAI ayaa soo saartay GPT-6 Sol iyo Luna, nidaamyadeeda cusub ee ku saabsan kharashka iyo kalsoonida. Nidaamyadan cusub waa kuwo aad u wanaagsan oo lacagta hooseeya, iyagoo diiradda saaray horumarro aad u yar laakiin muhiim u ah.
OpenAI ayaa sidoo kale soo saartay GPT-6 Astra, nidaamkeeda cusub ee ku saabsan kharashka iyo kalsoonida. Nidaamkan cusub waa mid aad u wanaagsan oo lacagta hooseeya, iyagoo diiradda saaray horumarro aad u yar laakiin muhiim u ah.
Muhiimadda horumarahan cusub
Horumarahan cusub waa kuwo aad u muhiim ah, iyagoo diiradda saaray kharashka iyo kalsoonida. Shirkadaha AI-ka ayaa diiradda saaray inay sameeyaan nidaamyo aad u wanaagsan oo lacagta hooseeya, iyagoo diiradda saaray horumarro aad u yar laakiin muhiim u ah.
Tani waxay muujisaa in shirkadaha AI-ka ay diiradda saarayaan inay sameeyaan nidaamyo aad u wanaagsan oo lacagta hooseeya, iyagoo diiradda saaray horumarro aad u yar laakiin muhiim u ah.
Anthropic and OpenAI have unveiled new AI models designed to deliver frontier-level performance at lower costs. Anthropic introduced Opus 5.5, while OpenAI released GPT-6 Sol and Luna, both targeting enterprise customers seeking speed and affordability.
Opus 5.5: A More Affordable Flagship
Anthropic has launched Opus 5.5, the latest iteration of its flagship model used for coding and complex knowledge tasks. The new version is priced at $4 per million input tokens and $20 per million output tokens, representing a 20% reduction compared to Opus 5.
Cache reads, which account for the majority of costs in agentic and coding workflows, are now $0.20 per million tokens, a 60% drop from the previous model. Additionally, Opus 5.5 generates output more than 30% faster than its predecessor.
OpenAI's GPT-6 Sol and Luna
OpenAI has introduced GPT-6 Sol and Luna, models focused on efficiency and speed. These are the latest versions of its mid-tier offerings, designed to compete with Anthropic's new release in terms of cost-effectiveness.
Earlier this month, OpenAI also released GPT-6 Astra, which has occasionally outperformed Opus 5 in benchmarks and user sentiment. However, Anthropic's new benchmarks show Opus 5.5 performing better in coding and knowledge work, albeit modestly.
The Race for Efficiency
Both companies argue that these releases push the frontier while substantially reducing costs. The focus is not on groundbreaking capabilities but on making frontier models more accessible to enterprise customers.
As organizations explore using model routers to favor cheaper alternatives, Anthropic and OpenAI are racing to maintain their competitive edge by offering improved efficiency and lower prices.
Ilaha iyo xuquuqda sawirka
Sawir: Ars Technica Xigasho
