DeepSeek·released 2026-04-221 source
Not yet benchmarked
Claims drawn from cited facts, not live model generation.
This model comes from DeepSeek and, at 290.94 billion parameters, sits firmly at the large end of the scale, having been released in 2026. That large scale buys considerably more capability headroom, but it comes at the cost of the efficiency and lower expense that a more compact model would offer.
4 cited facts
This model is not yet benchmarked: it is scored on no tracked benchmarks and holds no current top score.
2 cited facts
Its average gap to the state of the art is not disclosed, so no points-behind figure can be quoted here. With a leader count of zero, it holds the top score on none of its benchmarks, meaning there is no category leadership established yet.
2 cited facts
2 facts cross-checked across data sources: 2 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: huggingface_models
DeepSeek-V4-Flash is an AI model developed by DeepSeek, released 2026-04-22. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.