deepseek-v4.1VIDEO

DeepSeek V4.1 Flash 1017 hits 244 tok/s, faster than GPT-6.1 Sol Ultrafast

Reports V4.1 Flash 1017 generating 244 tokens per second, beating the $500/month GPT-6.1 Sol Ultrafast mode, plus Search & Fetch API free in Harness until Oct 31 then $0.75 per 1,000 calls.

David Arnal

@davidarngarArchived 11 days agoOriginally posted 2026.10.17

David Arnal @davidarngarOriginally posted 2026.10.17

Open original on X

DeepSeek V4.1 Flash 1017 generates 244 tokens per second, beating even GPT-6.1 Sol in Ultrafast mode, which costs $500/month. DeepSeek has also opened API access to its web search and page-fetching tools. Free through the official DeepSeek Harness until October 31. After that: $0.75 per 1,000 requests.

x.com/davidarngar/status/2111486571643797639Archived 2026.10.18 16:35