Model intelligence / Live profile
DeepSeek-R1
Track DeepSeek-R1 from Anthropic across launches, availability, pricing, benchmarks, capabilities, and consequential updates. This page updates automatically as substantive Superpower Daily coverage is published.
Coverage telemetryActive
Developer / company
AnthropicPublished9
Launches1
Pricing1
Benchmarks1
Published coverage9 stories
Last 90 days9 updates
Coverage dates7 days
Profile statusIndexed
ReleaseDeepSeek-V4.1-Flash
AvailabilityDeepSeek first-party API; US/API availability reported; also supported by Baseten and Ollama
Context1M
Input / 1M$0.3
Output / 1M$1.2
Latest coverage
What changed around DeepSeek-R1.
Durable signals
Launches, pricing, and proof.
011 records
Launches and releases
021 records
Pricing and access
031 records
Benchmarks and evaluations
Update timeline
The maintained record.
Pricing changeArs Technica Publishes Unitree Profile Citing Reliability and Retention ConcernsThe new profile connects founder Wang Xingxing’s hands-on cost discipline to low robot prices, while employee accounts raise questions about product quality and whether the company can scale.FundingDeepSeek Reportedly Hires CITIC Securities for Potential Shanghai IPOThe reported adviser appointment gives the AI developer a procedural route toward public markets, but a potential STAR Market deal has no settled terms.LaunchDeepSeek Launches V4.1-Flash and Plans to Replace V4-Pro API TrafficThe company says its smallest model in a new architecture family improves speed, cost and total runtime—but its own testing is driving the planned V4-Pro handoff.Model launchesDeepSeek launches open-weight V4.1-Flash modelDeepSeek launched V4.1-Flash, a multimodal open-weight model with a causal encoder-decoder architecture, 1M-token context, MIT licensing, and separate 8B prefill and 16B decode active-parameter paths.ResearchArizona Study Finds Repeated Falsehoods Can Sway AI Models in Long ChatsThe controlled seven-model experiment found large gaps in resistance and self-correction, while showing that factual reliability can change as a conversation continues.
Your watchlist