DeepSeek · Released 2026-10-17
A new MIT-licensed V4.1 Flash weight snapshot keeps the configuration, image/text input and text output. The API retains deepseek-flash and its prices. Independent AA/max scores 45 and Pi/high scores 100% within the small suite; older AA39 and LiveBench results remain with the earlier model.
Original model explanation ↗45points
Intelligence Index
Rank 12 among included models
Test configuration: max
Source read · · v4.3.2
Each model uses its highest tested setting, with fallback conditions retained. Original K3 remains at max44 from its readable individual page; the aggregate page now lists K3.1 Pro separately. Scores use one evaluator. The index mainly evaluates English text tasks; examine Chinese and image capabilities separately.
Source evaluation & rankings ↗100%
pass@1 (%)
Rank 1 among included models
Test configuration: Pi 0.85.1 · high · 30 attempts
Source read · · core10-v1.3 · Harbor 0.21.0 · Pi 0.85.1
Compare the same Pi version, high effort label and task suite; all five scores are equal and no winner is inferred.
Source evaluation & rankings ↗Published independent-results comparison, not our model test
On the evening of October 19, Artificial Analysis lists DeepSeek V4.1 Flash 1017 separately at max effort with an Intelligence Index of 45.
Published sources and an independent small suite, not our model test
On October 17, DeepSeek illustrates Flash 1017 with a Forbidden City scene, a browser racing game and a Godot game.
Independent engineering suite · not all coding
Official source, not our inference test
Separate official model and search/fetch billing
Official engineering explained, not a capability test or our deployment
Why can the same model feel quick through one service and slow through another?
Official tool access and a suggested workflow, not our execution
Search and Fetch are separate tool APIs: Search returns web results for a query; Fetch retrieves a chosen page as Markdown.
Official demos · no case source obtained or interaction tested
The V4 family reaches the API with Pro and Flash offering different capability and resource trade-offs; older request names temporarily map to Flash.
Original release ↗The architecture and size stay unchanged while post-training and Responses API support advance; this update applies only to the Flash API, not Pro or the web model.
Original release ↗Pro reaches general availability in the app, web and API with Responses API support and more flexible thinking effort; its request name stays unchanged.
Original release ↗A separate experimental vision model adds image understanding to Flash; V4.1 Flash later succeeds it.
Original release ↗Native multimodal Flash is released; legacy Flash API names temporarily route to the new model while Pro remains available.
Original release ↗A minor update ships different weights across website/app/API under the same deepseek-flash name and model prices; separate search/fetch APIs charge by request.
Original release ↗