AI Observatory / Model Radar DeepSeek / deepseek/deepseek-chat

DeepSeek: DeepSeek V3

DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations...

163,840 Context
16,384 Max output
$0.32 / 1M Prompt price
2024-12-26 Created
01 / Snapshot

Pricing, context, modalities, and parameters.

Model Radar detail pages stay neutral and operator-readable: core metadata first, then workflow fit.

Provider DeepSeek Input modalities text
Output modalities text Prompt price $0.32 / 1M
Completion price $0.89 / 1M Request price N/A
Context length 163,840 Max completion tokens 16,384
Supported parameters frequency_penalty, logit_bias, max_tokens, min_p, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p
Best for deepseek/deepseek-chat

DeepSeek: DeepSeek V3

DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations...

Coding workflows High-volume usage tool-capable low-cost
03 / Colophon

Routes and exits.

Each model page stays simple: overview, compare, related models, then back to the public hub.