Submit
DeepSeek AIOpen-weight·66K context·671B params·DeepSeek License v1.0

DeepSeek V3deepseek/deepseek-v3

DeepSeek V3 is a 671B-parameter MoE language model with strong reasoning and coding performance at prices roughly 1/10 of frontier US models. Open-weights released under DeepSeek License v1.0; OpenAI-compatible official API available globally.

Cheapest blended:$0.48 / 1M tokenson DeepSeek · 2 providers listed

Pricing across providers

Sort by:
ProviderInput /1MOutput /1MBlended /1MLatency p50FormatFreshnessAction

2 hostings hidden — prices older than 60 days are pending re-verification.

Affiliate disclosure: We may earn a commission from qualified signups. Pricing independence is enforced at the data layer — see our Editorial Independence Policy.

Works with

Point any of these clients at a hosting's base URL — they all speak at least one of this model's endpoint protocols (OPENAI_COMPATIBLE).

Capabilities

  • reasoning
  • coding
  • tool_use
  • long_context
  • multilingual

Languages: en, zh, ja, ko, es, fr, de, ru

Benchmarks

  • MMLU5-shot · official · source
    88.7%
  • HumanEval0-shot pass@1 · official · source
    82.5%

Code samples

Example using DeepSeek — the cheapest hosting for this model as of last verification. Swap base_url and model to use a different provider from the matrix above.

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.deepseek.com/v1",
)

response = client.chat.completions.create(
    model="deepseek-chat",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

Technical specs

Context
66K
Max output
8K
Parameters
671B
Release
2024-12-26
Training cutoff
2024-07-01
License
DeepSeek License v1.0
Paper →Hugging Face →Official blog →

Similar models

Compare with

Frequently asked

How much does DeepSeek V3 cost?+
The cheapest public hosting is $0.48 per 1M blended tokens on DeepSeek. 2 total providers are listed above with per-input / per-output / cached pricing.
How do I access DeepSeek V3 from outside China?+
All hostings listed above support global access. The official API (e.g. api.deepseek.com, dashscope-intl.aliyuncs.com) accepts international credit cards and does not require a Chinese mobile number. For privacy-sensitive workloads, third-party aggregators like Together.ai host the model on US/EU infrastructure.
Is DeepSeek V3 open-source? Can I fine-tune it?+
Yes. DeepSeek V3 is open-weight under the DeepSeek License v1.0 license. Weights are available on Hugging Face for local inference, fine-tuning, and commercial use (see license for specific terms).
Is DeepSeek V3 OpenAI-compatible?+
Most listed hostings expose an OpenAI-compatible API, so you can point an existing openai SDK client at the Provider's base_url and use the Provider's model name. See the Code Samples above for a copy-pasteable example.
What's the maximum context window for DeepSeek V3?+
The model supports up to 65,536 tokens of context (input + output). Some hosted versions may impose a smaller limit — check the "Context" column in the pricing matrix for each provider.