DeepSeek
DeepSeek-V4-Flash is a DeepSeek API model with 1M-token context, 384K maximum output, tool calls, JSON output, and OpenAI ChatCompletions and Anthropic-compatible access.
DeepSeek: DeepSeek V4 Flash is a model from DeepSeek. This standardized model-evaluation pass uses the stored catalog profile as evidence: DeepSeek-V4-Flash is a DeepSeek API model with 1M-token context, 384K maximum output, tool calls, JSON output, and OpenAI ChatCompletions and Anthropic-compatible access.
The strongest available signals are catalog use-case guidance, 6 stated capabilities, 2 public technical links, pricing notes in the catalog row, 1 evidence family with sources. The current weighted score is 71/100, led by technical capability at 100/100, coding and agentic work at 60/100, and developer experience at 88/100.
The main caveats are some evidence searches failed and are recorded as warnings. Treat this pending evaluation as a Kainotomic expert draft that synthesizes available public evidence rather than a claim that we ran every benchmark ourselves.
