AI coding platform Codeium, a California-based firm launched in 2021, has introduced that it’s integrating DeepSeek-R1 and V3 into its Windsurf platform, making Cascade one of many preliminary coding brokers to help R1.
The corporate mentioned that it might initially be priced at half the same old price and plans to cut back costs additional over time.
“R1 is really enjoyable and studying the chain of thought nearly appears like a requirement for reasoning fashions,” Codeium CEO Varun Mohan mentioned on X.
Final 12 months, Codeium built-in Anthropic’s Claude into Windsurf, a collaborative AI-native built-in growth setting (IDE).
Along with Codeium, Cursor introduced that the DeepSeek fashions can be found on its platform hosted on US servers.
Final 12 months, OpenAI’s newest o1 fashions have been made obtainable on Cursor. The o1 fashions have displayed distinctive efficiency in dealing with well-defined and complicated reasoning duties.
“Whereas we’re large followers of Deepseek, Sonnet nonetheless seems to carry out a lot better on real-world duties,” Cursor acknowledged in a put up on X.
Based by Michael Truell, Sualeh Asif, Arvid Lunnemark, and Aman Sanger, Cursor began with the objective of writing the world’s software program. Anysphere lately secured $100 million in a Sequence B funding spherical, bringing its post-money valuation to $2.6 billion.
Even Microsoft lately introduced it’s making DeepSeek-R1 obtainable on Azure AI Foundry and the GitHub Mannequin Catalogue, increasing the platform’s AI portfolio. “Clients will quickly be capable of run DeepSeek-R1’s distilled fashions regionally on Copilot+ PCs, in addition to on the huge ecosystem of GPUs obtainable on Home windows,” mentioned Microsoft chief Satya Nadella.
Even Amazon CEO Andy Jassy has introduced that the DeepSeek-R1 fashions at the moment are obtainable on Amazon Net Companies (AWS).
What makes DeepSeek Particular?
DeepSeek, a Chinese language AI analysis lab backed by Excessive-Flyer Capital Administration has launched DeepSeek-V3, the newest model of their frontier mannequin.
“The uncooked chain of thought from DeepSeek is fascinating. It actually reads like a human considering out loud: charming and unusual,” Ethan Mollick, professor at The Wharton College, mentioned.
Sharing related sentiments, Matthew Berman, CEO of Ahead Future, mentioned, “DeepSeek-R1 has essentially the most human-like inner monologue I’ve ever seen. It’s really fairly endearing.”
DeepSeek, in its analysis paper, revealed that the corporate wager large on reinforcement studying (RL) to coach each of those fashions. DeepSeek-R1-Zero was developed utilizing a pure RL method with none prior supervised fine-tuning (SFT). This mannequin utilised Group Relative Coverage Optimisation (GRPO), which permits for environment friendly RL coaching by estimating baselines from group scores moderately than requiring a separate critic mannequin of comparable dimension to the coverage mannequin.
DeepSeek-R1 incorporates a multi-stage coaching method and chilly begin knowledge. This technique improved the mannequin’s efficiency by refining its reasoning skills whereas sustaining readability in output. “The mannequin has proven efficiency corresponding to OpenAI’s o1-1217 on numerous reasoning duties,” the corporate mentioned.
“This ‘aha second’ within the DeepSeek-R1 paper is big. Pure reinforcement studying permits an LLM to mechanically be taught to assume and mirror,” Yuchen Jin, co-founder and CTO of Hyperbolic, mentioned.
The put up DeepSeek Involves Windsurf & Cursor appeared first on Analytics India Journal.
